[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-abhineet123--Deep-Learning-for-Tracking-and-Detection":3,"tool-abhineet123--Deep-Learning-for-Tracking-and-Detection":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",140436,2,"2026-04-05T23:32:43",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":80,"owner_email":78,"owner_twitter":78,"owner_website":81,"owner_url":82,"languages":83,"stars":92,"forks":93,"last_commit_at":94,"license":78,"difficulty_score":95,"env_os":96,"env_gpu":96,"env_ram":96,"env_deps":97,"category_tags":100,"github_topics":101,"view_count":112,"oss_zip_url":78,"oss_zip_packed_at":78,"status":16,"created_at":113,"updated_at":114,"faqs":115,"releases":121},119,"abhineet123\u002FDeep-Learning-for-Tracking-and-Detection","Deep-Learning-for-Tracking-and-Detection","Collection of papers, datasets, code and other resources for object tracking and detection using deep learning","Deep-Learning-for-Tracking-and-Detection 是一个面向目标检测与跟踪领域的开源资源集合，汇集了大量深度学习相关的论文、数据集、代码实现及其他实用资料。它系统整理了静态图像中的目标检测（如 YOLO、SSD、RetinaNet 等主流方法）、视频中的目标检测、单目标与多目标跟踪等多个方向的研究成果，并覆盖从传统方法到前沿技术（如无锚框检测、图神经网络、强化学习等）的广泛内容。该资源库特别适合计算机视觉领域的研究人员和开发者快速了解领域进展、复现经典算法或构建新模型。对于需要处理视频分析、自动驾驶、无人机监控、细胞追踪等任务的团队，也能从中找到对应的数据集和基准代码。其结构清晰、分类细致，是进入目标感知任务的理想起点。","\u003C!-- No Heading Fix -->\nCollection of papers, datasets, code and other resources for object detection and tracking using deep learning\n\u003C!-- MarkdownTOC -->\n\n- [Research Data](#research_data_)\n- [Papers](#paper_s_)\n    - [Static Detection](#static_detectio_n_)\n        - [Region Proposal](#region_proposal_)\n        - [RCNN](#rcn_n_)\n        - [YOLO](#yol_o_)\n        - [SSD](#ssd_)\n        - [RetinaNet](#retinanet_)\n        - [Anchor Free](#anchor_free_)\n        - [Misc](#mis_c_)\n    - [Video Detection](#video_detectio_n_)\n        - [Tubelet](#tubelet_)\n        - [FGFA](#fgf_a_)\n        - [RNN](#rnn_)\n    - [Multi Object Tracking](#multi_object_tracking_)\n        - [Joint-Detection](#joint_detection_)\n            - [Identity Embedding](#identity_embeddin_g_)\n        - [Association](#association_)\n        - [Deep Learning](#deep_learning_)\n        - [RNN](#rnn__1)\n        - [Unsupervised Learning](#unsupervised_learning_)\n        - [Reinforcement Learning](#reinforcement_learning_)\n        - [Network Flow](#network_flow_)\n        - [Graph Optimization](#graph_optimization_)\n        - [Baseline](#baselin_e_)\n        - [Metrics](#metrics_)\n    - [Single Object Tracking](#single_object_tracking_)\n        - [Reinforcement Learning](#reinforcement_learning__1)\n        - [Siamese](#siamese_)\n        - [Correlation](#correlation_)\n        - [Misc](#mis_c__1)\n    - [Deep Learning](#deep_learning__1)\n        - [Synthetic Gradients](#synthetic_gradient_s_)\n        - [Efficient](#efficient_)\n    - [Unsupervised Learning](#unsupervised_learning__1)\n    - [Interpolation](#interpolation_)\n    - [Autoencoder](#autoencoder_)\n        - [Variational](#variational_)\n- [Datasets](#dataset_s_)\n    - [Multi Object Tracking](#multi_object_tracking__1)\n        - [UAV](#uav_)\n        - [Synthetic](#synthetic_)\n        - [Microscopy \u002F Cell Tracking](#microscopy___cell_tracking_)\n    - [Single Object Tracking](#single_object_tracking__1)\n    - [Video Detection](#video_detectio_n__1)\n        - [Video Understanding \u002F Activity Recognition](#video_understanding___activity_recognitio_n_)\n    - [Static Detection](#static_detectio_n__1)\n        - [Animals](#animals_)\n    - [Boundary Detection](#boundary_detectio_n_)\n    - [Static Segmentation](#static_segmentation_)\n    - [Video Segmentation](#video_segmentation_)\n    - [Classification](#classificatio_n_)\n    - [Optical Flow](#optical_flow_)\n    - [Motion Prediction](#motion_prediction_)\n- [Code](#cod_e_)\n    - [General Vision](#general_vision_)\n    - [Multi Object Tracking](#multi_object_tracking__2)\n        - [Frameworks](#framework_s_)\n        - [General](#general_)\n        - [Baseline](#baselin_e__1)\n        - [Siamese](#siamese__1)\n        - [Unsupervised](#unsupervise_d_)\n        - [Re-ID](#re_id_)\n            - [Frameworks](#framework_s__1)\n        - [Graph NN](#graph_nn_)\n        - [Microscopy \u002F cell tracking](#microscopy___cell_tracking__1)\n        - [3D](#3_d_)\n        - [Metrics](#metrics__1)\n    - [Single Object Tracking](#single_object_tracking__2)\n        - [GUI Application \u002F Large Scale Tracking \u002F Animals](#gui_application___large_scale_tracking___animal_s_)\n    - [Video Detection](#video_detectio_n__2)\n        - [Action Detection](#action_detectio_n_)\n            - [Frameworks](#framework_s__2)\n    - [Static Detection and Matching](#static_detection_and_matching_)\n        - [Frameworks](#framework_s__3)\n        - [Region Proposal](#region_proposal__1)\n        - [FPN](#fpn_)\n        - [RCNN](#rcn_n__1)\n        - [SSD](#ssd__1)\n        - [RetinaNet](#retinanet__1)\n        - [YOLO](#yol_o__1)\n        - [Anchor Free](#anchor_free__1)\n        - [Misc](#mis_c__2)\n        - [Matching](#matchin_g_)\n        - [Boundary Detection](#boundary_detectio_n__1)\n        - [Text Detection](#text_detectio_n_)\n            - [Frameworks](#framework_s__4)\n        - [3D Detection](#3d_detectio_n_)\n            - [Frameworks](#framework_s__5)\n    - [Optical Flow](#optical_flow__1)\n        - [Frameworks](#framework_s__6)\n    - [Instance Segmentation](#instance_segmentation_)\n        - [Frameworks](#framework_s__7)\n    - [Semantic Segmentation](#semantic_segmentation_)\n        - [Frameworks](#framework_s__8)\n        - [Polyp](#polyp_)\n    - [Panoptic Segmentation](#panoptic_segmentation_)\n    - [Video Segmentation](#video_segmentation__1)\n        - [Panoptic Video Segmentation](#panoptic_video_segmentation_)\n    - [Motion Prediction](#motion_prediction__1)\n    - [Pose Estimation](#pose_estimation_)\n        - [Frameworks](#framework_s__9)\n    - [Autoencoders](#autoencoder_s_)\n    - [Classification](#classificatio_n__1)\n        - [Frameworks](#framework_s__10)\n    - [Deep RL](#deep_rl_)\n    - [Annotation](#annotatio_n_)\n        - [Editing](#editing_)\n        - [Augmentation](#augmentatio_n_)\n    - [Deep Learning](#deep_learning__2)\n        - [Class Imbalance](#class_imbalanc_e_)\n        - [Few shot learning](#few_shot_learning_)\n        - [Unsupervised learning](#unsupervised_learning__2)\n- [Collections](#collections_)\n    - [Datasets](#dataset_s__1)\n    - [Deep Learning](#deep_learning__3)\n    - [Static Detection](#static_detectio_n__2)\n    - [Video Detection](#video_detectio_n__3)\n    - [Single Object Tracking](#single_object_tracking__3)\n    - [Multi Object Tracking](#multi_object_tracking__3)\n    - [Static Segmentation](#static_segmentation__1)\n    - [Video Segmentation](#video_segmentation__2)\n    - [Motion Prediction](#motion_prediction__2)\n    - [Deep Compressed Sensing](#deep_compressed_sensin_g_)\n    - [Misc](#mis_c__3)\n- [Tutorials](#tutorials_)\n    - [Collections](#collections__1)\n    - [Multi Object Tracking](#multi_object_tracking__4)\n    - [Static Detection](#static_detectio_n__3)\n    - [Video Detection](#video_detectio_n__4)\n    - [Instance Segmentation](#instance_segmentation__1)\n    - [Deep Learning](#deep_learning__4)\n        - [Optimization](#optimizatio_n_)\n        - [Class Imbalance](#class_imbalanc_e__1)\n    - [RNN](#rnn__2)\n    - [Deep RL](#deep_rl__1)\n    - [Autoencoders](#autoencoder_s__1)\n- [Blogs](#blogs_)\n\n\u003C!-- \u002FMarkdownTOC -->\n\n\u003Ca id=\"research_data_\">\u003C\u002Fa>\n# Research Data\n\nI use [DavidRM Journal](http:\u002F\u002Fwww.davidrm.com\u002F) for managing my research data for its excellent hierarchical organization, cross-linking and tagging capabilities.\n\nI make available a Journal entry export file that contains tagged and categorized collection of papers, articles, tutorials, code and notes about computer vision and deep learning that I have collected over the last few years.\n\nThis is what the topic cloud looks like:\n![Alt text](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fabhineet123_Deep-Learning-for-Tracking-and-Detection_readme_439592ee4cb1.jpg)\n\nIt needs Jounal 8 and can be imported using following steps:\n \n- Import my user preferences using **File** -> **Import** -> **Import User Preferences**\n- Import research data using **File** -> **Import** -> **Sync from The Journal Export File**\n\nNote that my user preferences must be imported _before_ the research data for the tagged topics to work correctly.\n\n(optional) My global options file is also provided for those interested in a dark theme and can be imported using **File** -> **Import** -> **Import Global Options**\n\n-  [User Preferences](research_data\u002Fuser_settings.juser)\n-  [Entry Export File](research_data\u002Fphd_literature_readings.tjexp)\n-  [Global Options](research_data\u002Fglobal_options.tjglobal)\n\nUpdated: 2026-03-09\n\n\u003Ca id=\"paper_s_\">\u003C\u002Fa>\n# Papers\n\n\u003Ca id=\"static_detectio_n_\">\u003C\u002Fa>\n## Static Detection\n\n\u003Ca id=\"region_proposal_\">\u003C\u002Fa>\n### Region Proposal\n- **Scalable Object Detection Using Deep Neural Networks**\n[cvpr14]\n[[pdf]](static_detection\u002Fregion_proposal\u002FScalable%20Object%20Detection%20Using%20Deep%20Neural%20Networks%20cvpr14.pdf)\n[[notes]](static_detection\u002Fnotes\u002FScalable%20Object%20Detection%20Using%20Deep%20Neural%20Networks%20cvpr14.pdf)\n- **Selective Search for Object Recognition**\n[ijcv2013]\n[[pdf]](static_detection\u002Fregion_proposal\u002FSelective%20Search%20for%20Object%20Recognition%20ijcv2013.pdf)\n[[notes]](static_detection\u002Fnotes\u002FSelective%20Search%20for%20Object%20Recognition%20ijcv2013.pdf)\n\n\u003Ca id=\"rcn_n_\">\u003C\u002Fa>\n### RCNN\n- **Faster R-CNN Towards Real-Time Object Detection with Region Proposal Networks**\n[tpami17]\n[[pdf]](static_detection\u002FRCNN\u002FFaster%20R-CNN%20Towards%20Real-Time%20Object%20Detection%20with%20Region%20Proposal%20Networks%20tpami17%20ax16_1.pdf)\n[[notes]](static_detection\u002Fnotes\u002FFaster_R-CNN.pdf)\n- **RFCN - Object Detection via Region-based Fully Convolutional Networks**\n[nips16]\n[Microsoft Research]\n[[pdf]](static_detection\u002FRCNN\u002FRFCN-Object%20Detection%20via%20Region-based%20Fully%20Convolutional%20Networks%20nips16.pdf)\n[[notes]](static_detection\u002Fnotes\u002FRFCN.pdf)  \n- **Mask R-CNN**\n[iccv17]\n[Facebook AI Research]\n[[pdf]](static_detection\u002FRCNN\u002FMask%20R-CNN%20ax17_4%20iccv17.pdf)\n[[notes]](static_detection\u002Fnotes\u002FMask%20R-CNN%20ax17_4%20iccv17.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1703.06870)\n[[code (keras)]](https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)\n[[code (tensorflow)]](https:\u002F\u002Fgithub.com\u002FCharlesShang\u002FFastMaskRCNN)\n- **SNIPER Efficient Multi-Scale Training**\n[ax1812\u002Fnips18]\n[[pdf]](static_detection\u002FRCNN\u002FSNIPER%20Efficient%20Multi-Scale%20Training%20ax181213%20nips18.pdf)\n[[notes]](static_detection\u002Fnotes\u002FSNIPER%20Efficient%20Multi-Scale%20Training%20ax181213%20nips18.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fmahyarnajibi\u002FSNIPER)\n\n\n\u003Ca id=\"yol_o_\">\u003C\u002Fa>\n### YOLO\n- **You Only Look Once Unified, Real-Time Object Detection**\n[ax1605]\n[[pdf]](static_detection\u002Fyolo\u002FYou%20Only%20Look%20Once%20Unified,%20Real-Time%20Object%20Detection%20ax1605.pdf)\n[[notes]](static_detection\u002Fnotes\u002FYou%20Only%20Look%20Once%20Unified,%20Real-Time%20Object%20Detection%20ax1605.pdf)\n- **YOLO9000 Better, Faster, Stronger**\n[ax1612]\n[[pdf]](static_detection\u002Fyolo\u002FYOLO9000%20Better,%20Faster,%20Stronger%20ax16_12.pdf)\n[[notes]](static_detection\u002Fnotes\u002FYOLO9000%20Better,%20Faster,%20Stronger%20ax16_12.pdf)\n- **YOLOv3 An Incremental Improvement**\n[ax1804]\n[[pdf]](static_detection\u002Fyolo\u002FYOLOv3%20An%20Incremental%20Improvement%20ax180408.pdf)\n[[notes]](static_detection\u002Fnotes\u002FYOLOv3%20An%20Incremental%20Improvement%20ax180408.pdf)\n- **YOLOv4 Optimal Speed and Accuracy of Object Detection**\n[ax2004]\n[[pdf]](static_detection\u002Fyolo\u002FYOLOV4_Optimal%20Speed%20and%20Accuracy%20of%20Object%20Detection%20ax200423.pdf)\n[[notes]](static_detection\u002Fnotes\u002FYOLOV4_Optimal%20Speed%20and%20Accuracy%20of%20Object%20Detection%20ax200423.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002FAlexeyAB\u002Fdarknet)\n\n\u003Ca id=\"ssd_\">\u003C\u002Fa>\n### SSD\n- **SSD Single Shot MultiBox Detector**\n[ax1612\u002Feccv16]\n[[pdf]](static_detection\u002Fssd\u002FSSD%20Single%20Shot%20MultiBox%20Detector%20eccv16_ax16_12.pdf)\n[[notes]](static_detection\u002Fnotes\u002FSSD.pdf)\n- **DSSD  Deconvolutional Single Shot Detector**\n[ax1701]\n[[pdf]](static_detection\u002Fssd\u002FDSSD%20Deconvolutional%20Single%20Shot%20Detector%20ax1701.06659.pdf)\n[[notes]](static_detection\u002Fnotes\u002FDSSD.pdf)\n\n\u003Ca id=\"retinanet_\">\u003C\u002Fa>\n### RetinaNet\n- **Feature Pyramid Networks for Object Detection**\n[ax1704]\n[[pdf]](static_detection\u002Fretinanet\u002FFeature%20Pyramid%20Networks%20for%20Object%20Detection%20ax170419.pdf)\n[[notes]](static_detection\u002Fnotes\u002FFPN.pdf)\n- **Focal Loss for Dense Object Detection**\n[ax180207\u002Ficcv17]\n[[pdf]](static_detection\u002Fretinanet\u002FFocal%20Loss%20for%20Dense%20Object%20Detection%20ax180207%20iccv17.pdf)\n[[notes]](static_detection\u002Fnotes\u002Ffocal_loss.pdf) \n\n\u003Ca id=\"anchor_free_\">\u003C\u002Fa>\n### Anchor Free\n\n- **FoveaBox: Beyond Anchor-based Object Detector**\n[ax1904]\n[[pdf]](static_detection\u002Fanchor_free\u002FFoveaBox%20Beyond%20Anchor-based%20Object%20Detector%20ax1904.03797.pdf)\n[[notes]](static_detection\u002Fnotes\u002FFoveaBox%20Beyond%20Anchor-based%20Object%20Detector%20ax1904.03797.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Ftaokong\u002FFoveaBox)\n- **CornerNet: Detecting Objects as Paired Keypoints**\n[ax1903\u002Fijcv19]\n[[pdf]](static_detection\u002Fanchor_free\u002FCornerNet%20Detecting%20Objects%20as%20Paired%20Keypoints%20ax1903%20ijcv19.pdf)\n[[notes]](static_detection\u002Fnotes\u002FCornerNet%20Detecting%20Objects%20as%20Paired%20Keypoints%20ax1903%20ijcv19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fprinceton-vl\u002FCornerNet)\n- **FCOS Fully Convolutional One-Stage Object Detection**\n[ax1908\u002Ficcv19]\n[[pdf]](static_detection\u002Fanchor_free\u002FFCOS%20Fully%20Convolutional%20One-Stage%20Object%20Detection%20ax1908%20iccv19.pdf)\n[[notes]](static_detection\u002Fnotes\u002FFCOS%20Fully%20Convolutional%20One-Stage%20Object%20Detection%20ax1908%20iccv19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Ftianzhi0549\u002FFCOS)\n[[code\u002FFCOS_PLUS]](https:\u002F\u002Fgithub.com\u002Fyqyao\u002FFCOS_PLUS)\n[[code\u002FVoVNet]](https:\u002F\u002Fgithub.com\u002Fvov-net\u002FVoVNet-FCOS)\n[[code\u002FHRNet]](https:\u002F\u002Fgithub.com\u002FHRNet\u002FHRNet-FCOS)\n[[code\u002FNAS]](https:\u002F\u002Fgithub.com\u002FLausannen\u002FNAS-FCOS)\n- **Feature Selective Anchor-Free Module for Single-Shot Object Detection**\n[ax1903\u002Fcvpr19]\n[[pdf]](static_detection\u002Fanchor_free\u002FFeature%20Selective%20Anchor-Free%20Module%20for%20Single-Shot%20Object%20Detection%20ax1903.00621%20cvpr19.pdf)\n[[notes]](static_detection\u002Fnotes\u002FFeature%20Selective%20Anchor-Free%20Module%20for%20Single-Shot%20Object%20Detection%20ax1903.00621%20cvpr19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fhdjang\u002FFeature-Selective-Anchor-Free-Module-for-Single-Shot-Object-Detection)\n- **Bottom-up object detection by grouping extreme and center points**\n[ax1901]\n[[pdf]](static_detection\u002Fanchor_free\u002FBottom-up%20object%20detection%20by%20grouping%20extreme%20and%20center%20points%201901.08043.pdf)\n[[notes]](static_detection\u002Fnotes\u002FBottom-up%20object%20detection%20by%20grouping%20extreme%20and%20center%20points%201901.08043.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FExtremeNet)\n- **Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection**\n[ax1912\u002Fcvpr20]\n[[pdf]](static_detection\u002Fanchor_free\u002FBridging%20the%20Gap%20Between%20Anchor-based%20and%20Anchor-free%20Detection%20via%20Adaptive%20Training%20Sample%20Selection%201912.02424%20cvpr20.pdf)\n[[notes]](static_detection\u002Fnotes\u002FBridging%20the%20Gap%20Between%20Anchor-based%20and%20Anchor-free%20Detection%20via%20Adaptive%20Training%20Sample%20Selection%201912.02424%20cvpr20.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fsfzhang15\u002FATSS)\n- **End-to-end object detection with Transformers**\n[ax200528]\n[[pdf]](static_detection\u002Fanchor_free\u002FEnd-to-End%20Object%20Detection%20with%20Transformers%20ax200528.pdf)\n[[notes]](static_detection\u002Fnotes\u002FEnd-to-end%20object%20detection%20with%20Transformers%20ax200528.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetr)\n- **Objects as Points**\n[ax1904]\n[[pdf]](static_detection\u002Fanchor_free\u002FObjects%20as%20Points%20ax1904.07850.pdf)\n[[notes]](static_detection\u002Fnotes\u002FObjects%20as%20Points%20ax1904.07850.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterNet)\n- **RepPoints Point Set Representation for Object Detection**\n[iccv19]\n[[pdf]](static_detection\u002Fanchor_free\u002FRepPoints%20Point%20Set%20Representation%20for%20Object%20Detection%201904.11490%20iccv19.pdf)\n[[notes]](static_detection\u002Fnotes\u002FRepPoints%20Point%20Set%20Representation%20for%20Object%20Detection%201904.11490%20iccv19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FRepPoints)\n\n\u003Ca id=\"mis_c_\">\u003C\u002Fa>\n### Misc\n- **OverFeat Integrated Recognition, Localization and Detection using Convolutional Networks**\n[ax1402\u002Ficlr14]\n[[pdf]](static_detection\u002FOverFeat%20Integrated%20Recognition,%20Localization%20and%20Detection%20using%20Convolutional%20Networks%20ax1402%20iclr14.pdf)\n[[notes]](static_detection\u002Fnotes\u002FOverFeat%20Integrated%20Recognition,%20Localization%20and%20Detection%20using%20Convolutional%20Networks%20ax1402%20iclr14.pdf)\n- **LSDA Large scale detection through adaptation**\n[ax1411\u002Fnips14]\n[[pdf]](static_detection\u002FLSDA%20Large%20scale%20detection%20through%20adaptation%20nips14%20ax14_11.pdf)\n[[notes]](static_detection\u002Fnotes\u002FLSDA%20Large%20scale%20detection%20through%20adaptation%20nips14%20ax14_11.pdf)\n- **Acquisition of Localization Confidence for Accurate Object Detection**\n[ax1807\u002Feccv18]\n[[pdf]](static_detection\u002FAcquisition%20of%20Localization%20Confidence%20for%20Accurate%20Object%20Detection%201807.11590%20eccv18.pdf)\n[[notes]](static_detection\u002Fnotes\u002FIOU-Net.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fvacancy\u002FPreciseRoIPooling)\n- **EfficientDet: Scalable and Efficient Object Detection**\n[cvpr20]\n[[pdf]](static_detection\u002FEfficientDet_Scalable%20and%20efficient%20object%20detection.pdf)\n - **Generalized Intersection over Union A Metric and A Loss for Bounding Box Regression**\n[ax1902\u002Fcvpr19]\n[[pdf]](static_detection\u002FGeneralized%20Intersection%20over%20Union%20A%20Metric%20and%20A%20Loss%20for%20Bounding%20Box%20Regression%201902.09630%20cvpr19.pdf)\n[[notes]](static_detection\u002Fnotes\u002FGeneralized%20Intersection%20over%20Union%20A%20Metric%20and%20A%20Loss%20for%20Bounding%20Box%20Regression%201902.09630%20cvpr19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fgeneralized-iou)\n[[project]](https:\u002F\u002Fgiou.stanford.edu\u002F)\n\n\u003Ca id=\"video_detectio_n_\">\u003C\u002Fa>\n## Video Detection\n\n\u003Ca id=\"tubelet_\">\u003C\u002Fa>\n### Tubelet\n* **Object Detection from Video Tubelets with Convolutional Neural Networks**\n[cvpr16]\n[[pdf]](video_detection\u002Ftubelets\u002FObject_Detection_from_Video_Tubelets_with_Convolutional_Neural_Networks_CVPR16.pdf)\n[[notes]](video_detection\u002Fnotes\u002FObject_Detection_from_Video_Tubelets_with_Convolutional_Neural_Networks_CVPR16.pdf)\n* **Object Detection in Videos with Tubelet Proposal Networks**\n[ax1704\u002Fcvpr17]\n[[pdf]](video_detection\u002Ftubelets\u002FObject_Detection_in_Videos_with_Tubelet_Proposal_Networks_ax1704_cvpr17.pdf)\n[[notes]](video_detection\u002Fnotes\u002FObject_Detection_in_Videos_with_Tubelet_Proposal_Networks_ax1704_cvpr17.pdf)\n\n\u003Ca id=\"fgf_a_\">\u003C\u002Fa>\n### FGFA\n* **Deep Feature Flow for Video Recognition**\n[cvpr17]\n[Microsoft Research]\n[[pdf]](video_detection\u002Ffgfa\u002FDeep%20Feature%20Flow%20For%20Video%20Recognition%20cvpr17.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1611.07715)\n[[code]](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FDeep-Feature-Flow)   \n* **Flow-Guided Feature Aggregation for Video Object Detection**\n[ax1708\u002Ficcv17]\n[[pdf]](video_detection\u002Ffgfa\u002FFlow-Guided%20Feature%20Aggregation%20for%20Video%20Object%20Detection%20ax1708%20iccv17.pdf)\n[[notes]](video_detection\u002Fnotes\u002FFlow-Guided%20Feature%20Aggregation%20for%20Video%20Object%20Detection%20ax1708%20iccv17.pdf)\n* **Towards High Performance Video Object Detection**\n[ax1711]\n[Microsoft]\n[[pdf]](video_detection\u002Ffgfa\u002FTowards%20High%20Performance%20Video%20Object%20Detection%20ax171130%20microsoft.pdf)\n[[notes]](video_detection\u002Fnotes\u002FTowards%20High%20Performance%20Video%20Object%20Detection%20ax171130%20microsoft.pdf)\n\n\u003Ca id=\"rnn_\">\u003C\u002Fa>\n### RNN\n* **Online Video Object Detection using Association LSTM**\n[iccv17]\n[[pdf]](video_detection\u002Frnn\u002FOnline%20Video%20Object%20Detection%20using%20Association%20LSTM%20iccv17.pdf)\n[[notes]](video_detection\u002Fnotes\u002FOnline%20Video%20Object%20Detection%20using%20Association%20LSTM%20iccv17.pdf)\n* **Context Matters Refining Object Detection in Video with Recurrent Neural Networks**\n[bmvc16]\n[[pdf]](video_detection\u002Frnn\u002FContext%20Matters%20Refining%20Object%20Detection%20in%20Video%20with%20Recurrent%20Neural%20Networks%20bmvc16.pdf)\n[[notes]](video_detection\u002Fnotes\u002FContext%20Matters%20Refining%20Object%20Detection%20in%20Video%20with%20Recurrent%20Neural%20Networks%20bmvc16.pdf)\n\n\u003Ca id=\"multi_object_tracking_\">\u003C\u002Fa>\n##  Multi Object Tracking\n\n\u003Ca id=\"joint_detection_\">\u003C\u002Fa>\n### Joint-Detection\n\n* **Tracking Objects as Points**\n[ax2004]\n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FTracking%20Objects%20as%20Points%202004.01177.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking%20Objects%20as%20Points%202004.01177.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterTrack)[pytorch]\n\n\n\u003Ca id=\"identity_embeddin_g_\">\u003C\u002Fa>\n#### Identity Embedding \n\n* **MOTS Multi-Object Tracking and Segmentation**\n[cvpr19]\n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FMOTS%20Multi-Object%20Tracking%20and%20Segmentation%20ax1904%20cvpr19.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FMOTS%20Multi-Object%20Tracking%20and%20Segmentation%20ax1904%20cvpr19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FTrackR-CNN)\n[[project\u002Fdata]](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fmots)\n* **Towards Real-Time Multi-Object Tracking**\n[ax1909]\n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FTowards%20Real-Time%20Multi-Object%20Tracking%20ax1909.12605v1.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FTowards%20Real-Time%20Multi-Object%20Tracking%20ax1909.12605v1.pdf)\n* **A Simple Baseline for Multi-Object Tracking**\n[ax2004]\n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FA%20Simple%20Baseline%20for%20Multi-Object%20Tracking%202004.01888.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FA%20Simple%20Baseline%20for%20Multi-Object%20Tracking%202004.01888.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fifzhang\u002FFairMOT)\n\n* **Integrated Object Detection and Tracking with Tracklet-Conditioned Detection**\n[ax1811]\n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FIntegrated%20Object%20Detection%20and%20Tracking%20with%20Tracklet-Conditioned%20Detection%201811.11167.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FIntegrated%20Object%20Detection%20and%20Tracking%20with%20Tracklet-Conditioned%20Detection%201811.11167.pdf)\n\n\n\n\u003Ca id=\"association_\">\u003C\u002Fa>\n### Association\n\n* **Deep Affinity Network for Multiple Object Tracking**\n[ax1810\u002Ftpami19]\n[[pdf]](multi_object_tracking\u002Fassociation\u002FDeep%20Affinity%20Network%20for%20Multiple%20Object%20Tracking%20ax1810.11780%20tpami19.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FDeep%20Affinity%20Network%20for%20Multiple%20Object%20Tracking%20ax1810.11780%20tpami19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002FshijieS\u002FSST) [pytorch]\n\n\u003Ca id=\"deep_learning_\">\u003C\u002Fa>\n### Deep Learning\n\n* **Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal Attention Mechanism**\n[ax1708\u002Ficcv17]\n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FOnline%20Multi-Object%20Tracking%20Using%20CNN-based%20Single%20Object%20Tracker%20with%20Spatial-Temporal%20Attention%20Mechanism%201708.02843%20iccv17.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1708.02843)\n[[notes]](multi_object_tracking\u002Fnotes\u002FOnline%20Multi-Object%20Tracking%20Using%20CNN-based%20Single%20Object%20Tracker%20with%20Spatial-Temporal%20Attention%20Mechanism%201708.02843%20iccv17.pdf)\n* **Online multi-object tracking with dual matching attention networks**\n[ax1902\u002Feccv18]\n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FOnline%20multi-object%20tracking%20with%20dual%20matching%20attention%20networks%201902.00749%20eccv18.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1902.00749)\n[[notes]](multi_object_tracking\u002Fnotes\u002FOnline%20multi-object%20tracking%20with%20dual%20matching%20attention%20networks%201902.00749%20eccv18.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fjizhu1023\u002FDMAN_MOT)\n* **FAMNet Joint Learning of Feature, Affinity and Multi-Dimensional Assignment for Online Multiple Object Tracking**\n[iccv19]\n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FFAMNet%20Joint%20Learning%20of%20Feature,%20Affinity%20and%20Multi-Dimensional%20Assignment%20for%20Online%20Multiple%20Object%20Tracking%20iccv19.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FFAMNet%20Joint%20Learning%20of%20Feature,%20Affinity%20and%20Multi-Dimensional%20Assignment%20for%20Online%20Multiple%20Object%20Tracking%20iccv19.pdf)\n\n* **Exploit the Connectivity: Multi-Object Tracking with TrackletNet**\n[ax1811\u002Fmm19]\n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FExploit%20the%20Connectivity%20Multi-Object%20Tracking%20with%20TrackletNet%20ax1811.07258%20mm19.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FExploit%20the%20Connectivity%20Multi-Object%20Tracking%20with%20TrackletNet%20ax1811.07258%20mm19.pdf)\n* **Tracking without bells and whistles**\n[ax1903\u002Ficcv19]\n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FTracking%20without%20bells%20and%20whistles%20ax1903.05625%20iccv19.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking%20without%20bells%20and%20whistles%20ax1903.05625%20iccv19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fphil-bergmann\u002Ftracking_wo_bnw) [pytorch]\n\n\u003Ca id=\"rnn__1\">\u003C\u002Fa>\n### RNN\n* **Tracking The Untrackable: Learning To Track Multiple Cues with Long-Term Dependencies**\n[ax1704\u002Ficcv17]\n[Stanford]\n[[pdf]](multi_object_tracking\u002Frnn\u002FTracking%20The%20Untrackable%20Learning%20To%20Track%20Multiple%20Cues%20with%20Long-Term%20Dependencies%20ax17_4_iccv17.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking_The_Untrackable_Learning_To_Track_Multiple_Cues_with_Long-Term_Dependencies.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1701.01909)\n[[project]](http:\u002F\u002Fweb.stanford.edu\u002F~alahi\u002F),\n* **Multi-object Tracking with Neural Gating Using Bilinear LSTM**\n[eccv18]\n[[pdf]](multi_object_tracking\u002Frnn\u002FMulti-object%20Tracking%20with%20Neural%20Gating%20Using%20Bilinear%20LSTM_eccv18.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FMulti-object%20Tracking%20with%20Neural%20Gating%20Using%20Bilinear%20LSTM_eccv18.pdf)\n* **Eliminating Exposure Bias and Metric Mismatch in Multiple Object Tracking**\n[cvpr19]\n[[pdf]](multi_object_tracking\u002Frnn\u002FEliminating%20Exposure%20Bias%20and%20Metric%20Mismatch%20in%20Multiple%20Object%20Tracking%20cvpr19.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FEliminating%20Exposure%20Bias%20and%20Metric%20Mismatch%20in%20Multiple%20Object%20Tracking%20cvpr19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fmaksay\u002Fseq-train)\n\n\u003Ca id=\"unsupervised_learning_\">\u003C\u002Fa>\n### Unsupervised Learning\n* **Unsupervised Person Re-identification by Deep Learning Tracklet Association**\n[ax1809\u002Feccv18]\n[[pdf]](multi_object_tracking\u002Funsupervised\u002FUnsupervised%20Person%20Re-identification%20by%20Deep%20Learning%20Tracklet%20Association%201809.02874%20eccv18.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FUnsupervised%20Person%20Re-identification%20by%20Deep%20Learning%20Tracklet%20Association%201809.02874%20eccv18.pdf)\n* **Tracking by Animation: Unsupervised Learning of Multi-Object Attentive Trackers**\n[ax1809\u002Fcvpr19]\n[[pdf]](multi_object_tracking\u002Funsupervised\u002FTracking%20by%20Animation%20Unsupervised%20Learning%20of%20Multi-Object%20Attentive%20Trackers%20cvpr19%20ax1809.03137.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1809.03137)\n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking%20by%20Animation%20Unsupervised%20Learning%20of%20Multi-Object%20Attentive%20Trackers%20cvpr19%20ax1809.03137.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fzhen-he\u002Ftracking-by-animation)\n* **Simple Unsupervised Multi-Object Tracking**\n[ax2006]\n[[pdf]](multi_object_tracking\u002Funsupervised\u002FSimple%20Unsupervised%20Multi-Object%20Tracking%202006.02609.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FSimple%20Unsupervised%20Multi-Object%20Tracking%202006.02609.pdf)\n\n\u003Ca id=\"reinforcement_learning_\">\u003C\u002Fa>\n### Reinforcement Learning\n* **Learning to Track: Online Multi-object Tracking by Decision Making**\n[iccv15]\n[Stanford]\n[[pdf]](multi_object_tracking\u002Frl\u002FLearning%20to%20Track%20Online%20Multi-object%20Tracking%20by%20Decision%20Making%20%20iccv15.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FLearning_to_Track_Online_Multi-object_Tracking_by_Decision_Making__iccv15.pdf)\n[[code (matlab)]](https:\u002F\u002Fgithub.com\u002Fyuxng\u002FMDP_Tracking)\n[[project]](https:\u002F\u002Fyuxng.github.io\u002F)\n* **Collaborative Deep Reinforcement Learning for Multi-Object Tracking**\n[eccv18]\n[[pdf]](multi_object_tracking\u002Frl\u002FCollaborative%20Deep%20Reinforcement%20Learning%20for%20Multi-Object%20Tracking_eccv18.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FCollaborative%20Deep%20Reinforcement%20Learning%20for%20Multi-Object%20Tracking_eccv18.pdf)\n\n\u003Ca id=\"network_flow_\">\u003C\u002Fa>\n### Network Flow\n* **Near-Online Multi-target Tracking with Aggregated Local Flow Descriptor**\n[iccv15]\n[NEC Labs]\n[[pdf]](multi_object_tracking\u002Fnetwork_flow\u002FNear-online%20multi-target%20tracking%20with%20aggregated%20local%20%EF%AC%82ow%20descriptor%20iccv15.pdf)\n[[author]](http:\u002F\u002Fwww-personal.umich.edu\u002F~wgchoi\u002F)\n[[notes]](multi_object_tracking\u002Fnotes\u002FNOMT.pdf)  \n* **Deep Network Flow for Multi-Object Tracking**\n[cvpr17]\n[NEC Labs]\n[[pdf]](multi_object_tracking\u002Fnetwork_flow\u002FDeep%20Network%20Flow%20for%20Multi-Object%20Tracking%20cvpr17.pdf)\n[[supplementary]](multi_object_tracking\u002Fnetwork_flow\u002FDeep%20Network%20Flow%20for%20Multi-Object%20Tracking%20cvpr17_supplemental.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FDeep%20Network%20Flow%20for%20Multi-Object%20Tracking%20cvpr17.pdf)  \n* **Learning a Neural Solver for Multiple Object Tracking**\n[ax1912\u002Fcvpr20]\n[[pdf]](multi_object_tracking\u002Fnetwork_flow\u002FLearning%20a%20Neural%20Solver%20for%20Multiple%20Object%20Tracking%201912.07515%20cvpr20.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FLearning%20a%20Neural%20Solver%20for%20Multiple%20Object%20Tracking%201912.07515%20cvpr20.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fdvl-tum\u002Fmot_neural_solver)\n\n\u003Ca id=\"graph_optimization_\">\u003C\u002Fa>\n### Graph Optimization\n* **A Multi-cut Formulation for Joint Segmentation and Tracking of Multiple Objects**\n[ax1607]\n[highest MT on MOT2015]\n[University of Freiburg, Germany]\n[[pdf]](multi_object_tracking\u002Fbatch\u002FA%20Multi-cut%20Formulation%20for%20Joint%20Segmentation%20and%20Tracking%20of%20Multiple%20Objects%20ax16_9%20%5Bbest%20MT%20on%20MOT15%5D.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1607.06317)\n[[author]](https:\u002F\u002Flmb.informatik.uni-freiburg.de\u002Fpeople\u002Fkeuper\u002Fpublications.html)\n[[notes]](multi_object_tracking\u002Fnotes\u002FA_Multi-cut_Formulation_for_Joint_Segmentation_and_Tracking_of_Multiple_Objects.pdf)\n\n\u003Ca id=\"baselin_e_\">\u003C\u002Fa>\n### Baseline\n* **Simple Online and Realtime Tracking**\n[icip16]\n[[pdf]](multi_object_tracking\u002Fbaseline\u002FSimple%20Online%20and%20Realtime%20Tracking%20ax1707%20icip16.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FSimple%20Online%20and%20Realtime%20Tracking%20ax1707%20icip16.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fabewley\u002Fsort)\n* **High-Speed Tracking-by-Detection Without Using Image Information**\n[avss17]\n[[pdf]](multi_object_tracking\u002Fbaseline\u002FHigh-Speed%20Tracking-by-Detection%20Without%20Using%20Image%20Information%20avss17.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FHigh-Speed%20Tracking-by-Detection%20Without%20Using%20Image%20Information%20avss17.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fbochinski\u002Fiou-tracker)\n     \n\u003Ca id=\"metrics_\">\u003C\u002Fa>\n### Metrics\n* **HOTA A Higher Order Metric for Evaluating Multi-object Tracking**\n[ijcv20\u002F08]\n[[pdf]](multi_object_tracking\u002Fmetrics\u002FHOTA%20A%20Higher%20Order%20Metric%20for%20Evaluating%20Multi-object%20Tracking%20sl_open_2010_ijcv2008.pdf)\n[[notes]](multi_object_tracking\u002Fnotes\u002FHOTA%20A%20Higher%20Order%20Metric%20for%20Evaluating%20Multi-object%20Tracking%20sl_open_2010_ijcv2008.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002FJonathonLuiten\u002FHOTA-metrics)\n\n\u003Ca id=\"single_object_tracking_\">\u003C\u002Fa>\n## Single Object Tracking\n\n\u003Ca id=\"reinforcement_learning__1\">\u003C\u002Fa>\n### Reinforcement Learning\n* **Deep Reinforcement Learning for Visual Object Tracking in Videos**\n[ax1704] [USC-Santa Barbara, Samsung Research]\n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FDeep%20Reinforcement%20Learning%20for%20Visual%20Object%20Tracking%20in%20Videos%20ax17_4.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1701.08936)\n[[author]](http:\u002F\u002Fwww.cs.ucsb.edu\u002F~dazhang\u002F)\n[[notes]](single_object_tracking\u002Fnotes\u002FDeep_Reinforcement_Learning_for_Visual_Object_Tracking_in_Videos.pdf)  \n* **Visual Tracking by Reinforced Decision Making**\n[ax1702] [Seoul National University, Chung-Ang University]\n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FVisual%20Tracking%20by%20Reinforced%20Decision%20Making%20ax17_2.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1702.06291)\n[[author]](http:\u002F\u002Fcau.ac.kr\u002F~jskwon\u002F)\n[[notes]](single_object_tracking\u002Fnotes\u002FVisual_Tracking_by_Reinforced_Decision_Making_ax17.pdf)\n* **Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning**\n[cvpr17] [Seoul National University]\n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FAction-Decision%20Networks%20for%20Visual%20Tracking%20with%20Deep%20Reinforcement%20Learning%20%20cvpr17%20supplementary.pdf)\n[[supplementary]](single_object_tracking\u002Freinforcement_learning\u002FAction-Decision%20Networks%20for%20Visual%20Tracking%20with%20Deep%20Reinforcement%20Learning%20%20cvpr17.pdf)\n[[project]](https:\u002F\u002Fsites.google.com\u002Fview\u002Fcvpr2017-adnet)\n[[notes]](single_object_tracking\u002Fnotes\u002FAction-Decision_Networks_for_Visual_Tracking_with_Deep_Reinforcement_Learning_cvpr17.pdf) \n[[code]](https:\u002F\u002Fgithub.com\u002Fildoonet\u002Ftf-adnet-tracking) \n* **End-to-end Active Object Tracking via Reinforcement Learning**\n[ax1705]\n[Peking University, Tencent AI Lab]\n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FEnd-to-end%20Active%20Object%20Tracking%20via%20Reinforcement%20Learning%20ax17_5.pdf)\n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1705.10561)\n\n\u003Ca id=\"siamese_\">\u003C\u002Fa>\n### Siamese\n* **Fully-Convolutional Siamese Networks for Object Tracking**\n[eccv16]\n[[pdf]](single_object_tracking\u002Fsiamese\u002FFully-Convolutional%20Siamese%20Networks%20for%20Object%20Tracking%20eccv16_9.pdf)\n[[project]](https:\u002F\u002Fwww.robots.ox.ac.uk\u002F~luca\u002Fsiamese-fc.html)\n[[notes]](single_object_tracking\u002Fnotes\u002FSiameseFC.pdf)  \n* **High Performance Visual Tracking with Siamese Region Proposal Network**\n[cvpr18]\n[[pdf]](single_object_tracking\u002Fsiamese\u002FHigh%20Performance%20Visual%20Tracking%20with%20Siamese%20Region%20Proposal%20Network_cvpr18.pdf)\n[[author]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~qwang\u002F)\n[[notes]](single_object_tracking\u002Fnotes\u002FHigh%20Performance%20Visual%20Tracking%20with%20Siamese%20Region%20Proposal%20Network_cvpr18.pdf)  \n* **Siam R-CNN Visual Tracking by Re-Detection**\n[cvpr20]\n[[pdf]](single_object_tracking\u002Fsiamese\u002FSiam%20R-CNN%20Visual%20Tracking%20by%20Re-Detection%201911.12836%20cvpr20.pdf)\n[[notes]](single_object_tracking\u002Fnotes\u002FSiam%20R-CNN%20Visual%20Tracking%20by%20Re-Detection%201911.12836%20cvpr20.pdf)\n[[project]](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fsiamrcnn)\n[[code]](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FSiamR-CNN)  \n\n\u003Ca id=\"correlation_\">\u003C\u002Fa>\n### Correlation\n* **ATOM Accurate Tracking by Overlap Maximization**\n[cvpr19]\n[[pdf]](single_object_tracking\u002Fcorrelation\u002FATOM%20Accurate%20Tracking%20by%20Overlap%20Maximization%20ax1811.07628%20cvpr19.pdf)\n[[notes]](single_object_tracking\u002Fnotes\u002FATOM%20Accurate%20Tracking%20by%20Overlap%20Maximization%20ax1811.07628%20cvpr19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking)\n* **DiMP Learning Discriminative Model Prediction for Tracking**\n[iccv19]\n[[pdf]](single_object_tracking\u002Fcorrelation\u002FDiMP%20Learning%20Discriminative%20Model%20Prediction%20for%20Tracking%20ax1904.07220%20iccv19.pdf)\n[[notes]](single_object_tracking\u002Fnotes\u002FDiMP%20Learning%20Discriminative%20Model%20Prediction%20for%20Tracking%20ax1904.07220%20iccv19.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking)\n* **D3S – A Discriminative Single Shot Segmentation Tracker**\n[cvpr20]\n[[pdf]](single_object_tracking\u002Fcorrelation\u002FD3S%20–%20A%20Discriminative%20Single%20Shot%20Segmentation%20Tracker%201911.08862v1%20cvpr20.pdf)\n[[notes]](single_object_tracking\u002Fnotes\u002FD3S%20–%20A%20Discriminative%20Single%20Shot%20Segmentation%20Tracker%201911.08862v1%20cvpr20.pdf)\n[[code]](https:\u002F\u002Fgithub.com\u002Falanlukezic\u002Fd3s)\n\n\u003Ca id=\"mis_c__1\">\u003C\u002Fa>\n### Misc\n\n* **Bridging the Gap Between Detection and Tracking A Unified Approach**\n[iccv19]\n[[pdf]](single_object_tracking\u002FBridging%20the%20Gap%20Between%20Detection%20and%20Tracking%20A%20Unified%20Approach%20iccv19.pdf)\n[[notes]](single_object_tracking\u002Fnotes\u002FBridging%20the%20Gap%20Between%20Detection%20and%20Tracking%20A%20Unified%20Approach%20iccv19.pdf)\n\n\u003Ca id=\"deep_learning__1\">\u003C\u002Fa>\n##  Deep Learning\n\n- **Do Deep Nets Really Need to be Deep**\n[nips14]\n[[pdf]](deep_learning\u002Ftheory\u002FDo%20Deep%20Nets%20Really%20Need%20to%20be%20Deep%20ax1410%20nips14.pdf)\n[[notes]](deep_learning\u002Fnotes\u002FDo%20Deep%20Nets%20Really%20Need%20to%20be%20Deep%20ax1410%20nips14.pdf)\n\n\u003Ca id=\"synthetic_gradient_s_\">\u003C\u002Fa>\n### Synthetic Gradients\n- **Decoupled Neural Interfaces using Synthetic Gradients**\n[ax1608]\n[[pdf]](deep_learning\u002Fsynthetic_gradients\u002FDecoupled%20Neural%20Interfaces%20using%20Synthetic%20Gradients%20ax1608.05343.pdf)\n[[notes]](deep_learning\u002Fnotes\u002FDecoupled%20Neural%20Interfaces%20using%20Synthetic%20Gradients%20ax1608.05343.pdf)    \n- **Understanding Synthetic Gradients and Decoupled Neural Interfaces**\n[ax1703]\n[[pdf]](deep_learning\u002Fsynthetic_gradients\u002FUnderstanding%20Synthetic%20Gradients%20and%20Decoupled%20Neural%20Interfaces%20ax1703.00522.pdf)\n[[notes]](deep_learning\u002Fnotes\u002FUnderstanding%20Synthetic%20Gradients%20and%20Decoupled%20Neural%20Interfaces%20ax1703.00522.pdf)\n\n\u003Ca id=\"efficient_\">\u003C\u002Fa>\n### Efficient\n- **EfficientNet:Rethinking Model Scaling for Convolutional Neural Networks**\n[icml2019]\n[[pdf]](deep_learning\u002Fefficient\u002FEfficientNet_Rethinking%20model%20scaling%20for%20CNNs.pdf)\n[[notes]](deep_learning\u002Fnotes\u002FEfficientNet_%20Rethinking%20Model%20Scaling%20for%20Convolutional%20Neural%20Networks.pdf)\n\n\u003Ca id=\"unsupervised_learning__1\">\u003C\u002Fa>\n## Unsupervised Learning\n- **Learning Features by Watching Objects Move**\n(cvpr17)\n[[pdf]](unsupervised\u002Fsegmentation\u002FLearning%20Features%20by%20Watching%20Objects%20Move%20ax170412%20cvpr17.pdf)\n[[notes]](unsupervised\u002Fnotes\u002FLearning%20Features%20by%20Watching%20Objects%20Move%20ax170412%20cvpr17.pdf)\n    \n\u003Ca id=\"interpolation_\">\u003C\u002Fa>\n##  Interpolation\n- **Video Frame Interpolation via Adaptive Convolution**\n[cvpr17 \u002F iccv17]\n[[pdf (cvpr17)]](interpolation\u002FVideo%20Frame%20Interpolation%20via%20Adaptive%20Convolution%20ax1703.pdf)\n[[pdf (iccv17)]](interpolation\u002FVideo%20Frame%20Interpolation%20via%20Adaptive%20Separable%20Convolution%20iccv17.pdf)\n[[ppt]](interpolation\u002Fnotes\u002FVideo%20Frame%20Interpolation%20via%20Adaptive%20Convolution%20ax1703.pdf)\n\n\u003Ca id=\"autoencoder_\">\u003C\u002Fa>\n## Autoencoder\n\n\u003Ca id=\"variational_\">\u003C\u002Fa>\n### Variational\n- **beta-VAE Learning Basic Visual Concepts with a Constrained Variational Framework** [iclr17]\n[[pdf]](autoencoder\u002Fvariational\u002Fbeta-VAE%20Learning%20Basic%20Visual%20Concepts%20with%20a%20Constrained%20Variational%20Framework%20iclr17.pdf)\n[[notes]](autoencoder\u002Fnotes\u002Fbeta-VAE%20Learning%20Basic%20Visual%20Concepts%20with%20a%20Constrained%20Variational%20Framework%20iclr17.pdf)\n- **Disentangling by Factorising** [ax1806]\n[[pdf]](autoencoder\u002Fvariational\u002FDisentangling%20by%20Factorising%20ax1806.pdf)\n[[notes]](autoencoder\u002Fnotes\u002FDisentangling%20by%20Factorising%20ax1806.pdf)  \n    \n\u003Ca id=\"dataset_s_\">\u003C\u002Fa>\n# Datasets\n\n\u003Ca id=\"multi_object_tracking__1\">\u003C\u002Fa>\n## Multi Object Tracking\n\n- [IDOT](https:\u002F\u002Fgithub.com\u002Fbitslab\u002FIDOT_dataset)\n- [UA-DETRAC Benchmark Suite](http:\u002F\u002Fdetrac-db.rit.albany.edu\u002F)\n- [GRAM Road-Traffic Monitoring](http:\u002F\u002Fagamenon.tsc.uah.es\u002FPersonales\u002Frlopez\u002Fdata\u002Frtm\u002F) [[paper]](https:\u002F\u002Flink.springer.com\u002Fchapter\u002F10.1007\u002F978-3-642-38622-0_32)\n- [Ko-PER Intersection Dataset](http:\u002F\u002Fwww.uni-ulm.de\u002Fin\u002Fmrm\u002Fforschung\u002Fdatensaetze.html)\n- [TRANCOS](http:\u002F\u002Fagamenon.tsc.uah.es\u002FPersonales\u002Frlopez\u002Fdata\u002Ftrancos\u002F)\n- [Urban Tracker](https:\u002F\u002Fwww.jpjodoin.com\u002Furbantracker\u002Fdataset.html)\n- [DARPA VIVID \u002F PETS 2005](http:\u002F\u002Fvision.cse.psu.edu\u002Fdata\u002FvividEval\u002Fdatasets\u002Fdatasets.html) [Non stationary camera]\n- [KIT-AKS](http:\u002F\u002Fi21www.ira.uka.de\u002Fimage_sequences\u002F) [No ground truth]\n- [CBCL StreetScenes Challenge Framework](http:\u002F\u002Fcbcl.mit.edu\u002Fsoftware-datasets\u002Fstreetscenes\u002F) [No top down viewpoint]\n- [MOT 2015](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002F2D_MOT_2015\u002F) [mostly street level viewpoint]\n- [MOT 2016](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FMOT16\u002F) [mostly street level viewpoint]\n- [MOT 2017](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FMOT17\u002F) [mostly street level viewpoint]\n- [MOT 2020](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FMOT20\u002F) [mostly top down  viewpoint]\n- [MOTS: Multi-Object Tracking and Segmentation](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fmots) [MOT and KITTI]\n- [CVPR 2019](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002F11) [mostly street level viewpoint]\n- [PETS 2009](http:\u002F\u002Fwww.cvg.reading.ac.uk\u002FPETS2009\u002Fa.html) [No vehicles]\n- [PETS 2017](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FPETS2017\u002F) [Low density] [mostly pedestrians]\n- [DukeMTMC](http:\u002F\u002Fvision.cs.duke.edu\u002FDukeMTMC\u002F) [multi camera] [static background] [pedestrians] [above-street level viewpoint] [website not working]\n- [KITTI Tracking Dataset](http:\u002F\u002Fwww.cvlibs.net\u002Fdatasets\u002Fkitti\u002Feval_tracking.php) [No top down viewpoint] [non stationary camera]\n- [The WILDTRACK Seven-Camera HD Dataset](https:\u002F\u002Fcvlab.epfl.ch\u002Fdata\u002Fdata-wildtrack\u002F) [pedestrian detection and tracking]\n- [3D Traffic Scene Understanding from Movable Platforms](http:\u002F\u002Fwww.cvlibs.net\u002Fprojects\u002Fintersection\u002F) [intersection traffic] [stereo setup] [moving camera]\n- [LOST : Longterm Observation of Scenes with Tracks](http:\u002F\u002Flost.cse.wustl.edu\u002F) [top down and street level viewpoint] [no ground truth]\n- [JTA](http:\u002F\u002Fimagelab.ing.unimore.it\u002Fimagelab\u002Fpage.asp?IdPage=25) [top down and street level viewpoint] [synthetic\u002FGTA 5] [pedestrian] [3D annotations]\n- [PathTrack: Fast Trajectory Annotation with Path Supervision](http:\u002F\u002Fpeople.ee.ethz.ch\u002F~daid\u002Fpathtrack\u002F) [top down and street level viewpoint] [iccv17] [pedestrian] \n- [CityFlow](https:\u002F\u002Fwww.aicitychallenge.org\u002F) [pole mounted] [intersections] [vehicles] [re-id]  [cvpr19]\n- [JackRabbot Dataset](https:\u002F\u002Fjrdb.stanford.edu\u002F)  [RGBD] [head-on][indoor\u002Foutdoor][stanford]\n- [TAO: A Large-Scale Benchmark for Tracking Any Object](http:\u002F\u002Ftaodataset.org\u002F)  [eccv20] [[code]](https:\u002F\u002Fgithub.com\u002FTAO-Dataset\u002Ftao)\n- [Edinburgh office monitoring video dataset](http:\u002F\u002Fhomepages.inf.ed.ac.uk\u002Frbf\u002FOFFICEDATA\u002F)  [indoors][long term][mostly static people]\n- [Waymo Open Dataset](https:\u002F\u002Fwaymo.com\u002Fopen\u002F)  [outdoors][vehicles]\n\n\u003Ca id=\"uav_\">\u003C\u002Fa>\n### UAV\n\n- [Stanford Drone Dataset](http:\u002F\u002Fcvgl.stanford.edu\u002Fprojects\u002Fuav_data\u002F)\n- [UAVDT - The Unmanned Aerial Vehicle Benchmark: Object Detection and Tracking](https:\u002F\u002Fsites.google.com\u002Fsite\u002Fdaviddo0323\u002Fprojects\u002Fuavdt) [uav] [intersections\u002Fhighways] [vehicles]  [eccv18]\n- [VisDrone](https:\u002F\u002Fgithub.com\u002FVisDrone\u002FVisDrone-Dataset) \n\n\u003Ca id=\"synthetic_\">\u003C\u002Fa>\n### Synthetic\n\n- [MNIST-MOT \u002F MNIST-Sprites ](https:\u002F\u002Fgithub.com\u002Fzhen-he\u002Ftracking-by-animation)  [script generated] [cvpr19]\n- [TUB Multi-Object and Multi-Camera Tracking Dataset ](https:\u002F\u002Fwww.nue.tu-berlin.de\u002Fmenue\u002Fforschung\u002Fsoftware_und_datensaetze\u002Fmocat\u002F)  [avss16]\n- [Virtual KITTI](http:\u002F\u002Fwww.xrce.xerox.com\u002FResearch-Development\u002FComputer-Vision\u002FProxy-Virtual-Worlds) [[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1605.06457)   [cvpr16] [link seems broken]\n \n\u003Ca id=\"microscopy___cell_tracking_\">\u003C\u002Fa>\n### Microscopy \u002F Cell Tracking\n\n- [Cell Tracking Challenge](http:\u002F\u002Fcelltrackingchallenge.net\u002F)  [nature methods\u002F2017]\n- [CTMC: Cell Tracking with Mitosis Detection Dataset Challenge ](https:\u002F\u002Fivc.ischool.utexas.edu\u002Fctmc\u002F)  [cvprw20] [[MOT]](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FCTMC-v1\u002F)\n \n\u003Ca id=\"single_object_tracking__1\">\u003C\u002Fa>\n## Single Object Tracking\n\n- [TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild](https:\u002F\u002Ftracking-net.org\u002F) [eccv18]\n- [LaSOT: Large-scale Single Object Tracking](https:\u002F\u002Fcis.temple.edu\u002Flasot\u002F) [cvpr19]\n- [Need for speed: A benchmark for higher frame rate object tracking](http:\u002F\u002Fci2cv.net\u002Fnfs\u002Findex.html) [iccv17]\n- [Long-term Tracking in the Wild A Benchmark](https:\u002F\u002Foxuva.github.io\u002Flong-term-tracking-benchmark\u002F) [eccv18]\n- [UAV123: A benchmark and simulator for UAV tracking](https:\u002F\u002Fuav123.org\u002F) [eccv16] [[project]](https:\u002F\u002Fivul.kaust.edu.sa\u002FPages\u002Fpub-benchmark-simulator-uav.aspx)\n- [Sim4CV A Photo-Realistic Simulator for Computer Vision Applications](https:\u002F\u002Fsim4cv.org\u002F) [ijcv18]   \n- [CDTB: A Color and Depth Visual Object Tracking and Benchmark](https:\u002F\u002Fwww.vicos.si\u002FProjects\u002FCDTB) [iccv19]   [RGBD]\n- [Temple Color 128 - Color Tracking Benchmark](http:\u002F\u002Fwww.dabi.temple.edu\u002F~hbling\u002Fdata\u002FTColor-128\u002FTColor-128.html) [tip15]\n\n\u003Ca id=\"video_detectio_n__1\">\u003C\u002Fa>\n## Video Detection\n\n- [YouTube-BB](https:\u002F\u002Fresearch.google.com\u002Fyoutube-bb\u002Fdownload.html)\n- [Imagenet-VID](http:\u002F\u002Fbvisionweb1.cs.unc.edu\u002Filsvrc2015\u002Fdownload-videos-3j16.php)\n\n\u003Ca id=\"video_understanding___activity_recognitio_n_\">\u003C\u002Fa>\n### Video Understanding \u002F Activity Recognition\n \n- [YouTube-8M](https:\u002F\u002Fresearch.google.com\u002Fyoutube8m\u002F)\n- [AVA: A Video Dataset of Atomic Visual Action](https:\u002F\u002Fresearch.google.com\u002Fava\u002F)\n- [VIRAT Video Dataset](http:\u002F\u002Fwww.viratdata.org\u002F)\n- [Kinetics Action Recognition Dataset](https:\u002F\u002Fdeepmind.com\u002Fresearch\u002Fopen-source\u002Fkinetics)\n\n\u003Ca id=\"static_detectio_n__1\">\u003C\u002Fa>\n## Static Detection\n- [PASCAL Visual Object Classes](http:\u002F\u002Fhost.robots.ox.ac.uk\u002Fpascal\u002FVOC\u002F)\n- [A Large-Scale Dataset for Vehicle Re-Identification in the Wild](https:\u002F\u002Fgithub.com\u002FPKU-IMRE\u002FVERI-Wild)\n[cvpr19]\n- [Object Detection-based annotations for some frames of the VIRAT dataset](https:\u002F\u002Fgithub.com\u002Fahrnbom\u002FViratAnnotationObjectDetection)\n- [MIO-TCD: A new benchmark dataset for vehicle classification and localization](http:\u002F\u002Fpodoce.dinf.usherbrooke.ca\u002Fchallenge\u002Fdataset\u002F) [tip18]\n- [Tiny ImageNet](https:\u002F\u002Ftiny-imagenet.herokuapp.com\u002F)\n \n\u003Ca id=\"animals_\">\u003C\u002Fa>\n### Animals\n\n- [Wildlife Image and Localization Dataset (species and bounding box labels)](https:\u002F\u002Flev.cs.rpi.edu\u002Fpublic\u002Fdatasets\u002Fwild.tar.gz)\n[wacv18]\n- [Stanford Dogs Dataset](http:\u002F\u002Fvision.stanford.edu\u002Faditya86\u002FImageNetDogs\u002F)\n[cvpr11]\n- [Oxford-IIIT Pet Dataset](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~vgg\u002Fdata\u002Fpets\u002F)\n[cvpr12]\n- [Caltech-UCSD Birds 200](http:\u002F\u002Fwww.vision.caltech.edu\u002Fvisipedia\u002FCUB-200.html) [rough segmentation] [attributes]\n- [Gold Standard Snapshot Serengeti Bounding Box Coordinates](https:\u002F\u002Fdataverse.scholarsportal.info\u002Fdataset.xhtml?persistentId=doi:10.5683\u002FSP\u002FTPB5ID)\n\n\u003Ca id=\"boundary_detectio_n_\">\u003C\u002Fa>\n## Boundary Detection\n\n- [Semantic Boundaries Dataset and Benchmark](http:\u002F\u002Fhome.bharathh.info\u002Fpubs\u002Fcodes\u002FSBD\u002Fdownload.html)\n\n\u003Ca id=\"static_segmentation_\">\u003C\u002Fa>\n## Static Segmentation\n\n- [COCO - Common Objects in Context](http:\u002F\u002Fcocodataset.org\u002F#download)\n- [Open Images](https:\u002F\u002Fstorage.googleapis.com\u002Fopenimages\u002Fweb\u002Findex.html)\n- [ADE20K](https:\u002F\u002Fgroups.csail.mit.edu\u002Fvision\u002Fdatasets\u002FADE20K\u002F) [cvpr17]\n- [SYNTHIA](http:\u002F\u002Fsynthia-dataset.net\u002Fdownload-2\u002F) [cvpr16]\n- [UC Berkeley Computer Vision Group - Contour Detection and Image Segmentation](https:\u002F\u002Fwww2.eecs.berkeley.edu\u002FResearch\u002FProjects\u002FCS\u002Fvision\u002Fgrouping\u002Fresources.html)\n\n\u003Ca id=\"video_segmentation_\">\u003C\u002Fa>\n## Video Segmentation\n\n- [DAVIS: Densely Annotated VIdeo Segmentation](https:\u002F\u002Fdavischallenge.org\u002F)\n- [Mapillary Vistas Dataset](https:\u002F\u002Fwww.mapillary.com\u002Fdataset\u002Fvistas?pKey=0_xJqX3-c-KyTb90oG_8HQ&lat=20&lng=0&z=1.5) [street scenes] [semi-free]\n- [BDD100K](https:\u002F\u002Fbair.berkeley.edu\u002Fblog\u002F2018\u002F05\u002F30\u002Fbdd\u002F) [street scenes] [autonomous driving]\n- [ApolloScape](http:\u002F\u002Fapolloscape.auto\u002F) [street scenes] [autonomous driving]\n- [Cityscapes](https:\u002F\u002Fwww.cityscapes-dataset.com\u002F) [street scenes] [instance-level]\n-  [YouTube-VOS](https:\u002F\u002Fyoutube-vos.org\u002Fdataset\u002Fvis\u002F) [iccv19]\n\n\u003Ca id=\"classificatio_n_\">\u003C\u002Fa>\n## Classification\n\n- [ImageNet Large Scale Visual Recognition Competition 2012](http:\u002F\u002Fwww.image-net.org\u002Fchallenges\u002FLSVRC\u002F2012\u002F)\n- [Animals with Attributes 2](https:\u002F\u002Fcvml.ist.ac.at\u002FAwA2\u002F)\n- [CompCars Dataset](http:\u002F\u002Fmmlab.ie.cuhk.edu.hk\u002Fdatasets\u002Fcomp_cars\u002Findex.html)\n- [ObjectNet](https:\u002F\u002Fobjectnet.dev\u002F) [only test set]\n\n\u003Ca id=\"optical_flow_\">\u003C\u002Fa>\n## Optical Flow\n\n- [Middlebury](http:\u002F\u002Fvision.middlebury.edu\u002Fflow\u002Fdata\u002F)\n- [MPI Sintel](http:\u002F\u002Fsintel.is.tue.mpg.de\u002F)\n- [KITTI Flow](http:\u002F\u002Fwww.cvlibs.net\u002Fdatasets\u002Fkitti\u002Feval_scene_flow.php?benchmark=flow)\n\n\u003Ca id=\"motion_prediction_\">\u003C\u002Fa>\n## Motion Prediction\n\n- [Trajnet++ (A Trajectory Forecasting Challenge)](https:\u002F\u002Fwww.aicrowd.com\u002Fchallenges\u002Ftrajnet-a-trajectory-forecasting-challenge)\n- [Trajectory Forecasting Challenge](http:\u002F\u002Ftrajnet.stanford.edu\u002F)\n\n\n\u003Ca id=\"cod_e_\">\u003C\u002Fa>\n# Code\n\n\u003Ca id=\"general_vision_\">\u003C\u002Fa>\n## General Vision\n- [Gluon CV Toolkit](https:\u002F\u002Fgithub.com\u002Fdmlc\u002Fgluon-cv) [mxnet] [pytorch]\n- [OpenMMLab Computer Vision Foundation](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmcv) [pytorch]\n\n\u003Ca id=\"multi_object_tracking__2\">\u003C\u002Fa>\n## Multi Object Tracking\n\n\u003Ca id=\"framework_s_\">\u003C\u002Fa>\n### Frameworks\n* [OpenMMLab Video Perception Toolbox. It supports Video Object Detection (VID), Multiple Object Tracking (MOT), Single Object Tracking (SOT), Video Instance Segmentation (VIS) with a unified framework](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmtracking) [pytorch]\n\n\u003Ca id=\"general_\">\u003C\u002Fa>\n### General\n* [Globally-optimal greedy algorithms for tracking a variable number of objects](http:\u002F\u002Fwww.csee.umbc.edu\u002F~hpirsiav\u002Fpapers\u002Ftracking_release_v1.0.tar.gz) [cvpr11] [matlab] [[author]](https:\u002F\u002Fwww.csee.umbc.edu\u002F~hpirsiav\u002F)    \n* [Continuous Energy Minimization for Multitarget Tracking](https:\u002F\u002Fbitbucket.org\u002Familan\u002Fcontracking) [cvpr11 \u002F iccv11 \u002F tpami  2014] [matlab]\n* [Discrete-Continuous Energy Minimization for Multi-Target Tracking](http:\u002F\u002Fwww.milanton.de\u002Ffiles\u002Fsoftware\u002Fdctracking-v1.0.zip) [cvpr12] [matlab] [[project]](http:\u002F\u002Fwww.milanton.de\u002Fdctracking\u002Findex.html)\n* [The way they move: Tracking multiple targets with similar appearance](https:\u002F\u002Fbitbucket.org\u002Fcdicle\u002Fsmot\u002Fsrc\u002Fmaster\u002F) [iccv13] [matlab]   \n* [3D Traffic Scene Understanding from Movable Platforms](http:\u002F\u002Fwww.cvlibs.net\u002Fprojects\u002Fintersection\u002F) [[2d_tracking]](http:\u002F\u002Fwww.cvlibs.net\u002Fsoftware\u002Ftrackbydet\u002F) [pami14\u002Fkit13\u002Ficcv13\u002Fnips11] [c++\u002Fmatlab]\n* [Multiple target tracking based on undirected hierarchical relation hypergraph](http:\u002F\u002Fwww.cbsr.ia.ac.cn\u002Fusers\u002Flywen\u002Fcodes\u002FMultiCarTracker.zip) [cvpr14] [C++] [[author]](http:\u002F\u002Fwww.cbsr.ia.ac.cn\u002Fusers\u002Flywen\u002F)\n* [Robust online multi-object tracking based on tracklet confidence and online discriminative appearance learning](https:\u002F\u002Fdrive.google.com\u002Fopen?id=1YMqvkrVI6LOXRwcaUlAZTu_b2_5GmTAM) [cvpr14] [matlab] [(project)](https:\u002F\u002Fsites.google.com\u002Fview\u002Finuvision\u002Fresearch)\n* [Learning to Track: Online Multi-Object Tracking by Decision Making](https:\u002F\u002Fgithub.com\u002Fyuxng\u002FMDP_Tracking) [iccv15] [matlab]\n* [Joint Tracking and Segmentation of Multiple Targets](https:\u002F\u002Fbitbucket.org\u002Familan\u002Fsegtracking) [cvpr15] [matlab]\n* [Multiple Hypothesis Tracking Revisited](http:\u002F\u002Frehg.org\u002Fmht\u002F) [iccv15] [highest MT on MOT2015 among open source trackers] [matlab]\n* [Combined Image- and World-Space Tracking in Traffic Scenes](https:\u002F\u002Fgithub.com\u002Faljosaosep\u002Fciwt) [icra 2017] [c++]\n* [Online Multi-Target Tracking with Recurrent Neural Networks](https:\u002F\u002Fbitbucket.org\u002Familan\u002Frnntracking\u002Fsrc\u002Fdefault\u002F) [aaai17] [lua\u002Ftorch7]\n* [Real-Time Multiple Object Tracking - A Study on the Importance of Speed](https:\u002F\u002Fgithub.com\u002Fsamuelmurray\u002Ftracking-by-detection) [ax1710\u002Fmasters thesis] [c++]        \n* [Beyond Pixels: Leveraging Geometry and Shape Cues for Online Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002FJunaidCS032\u002FMOTBeyondPixels) [icra18] [matlab]    \n* [Online Multi-Object Tracking with Dual Matching Attention Network](https:\u002F\u002Fgithub.com\u002Fjizhu1023\u002FDMAN_MOT) [eccv18] [matlab\u002Ftensorflow]    \n* [TrackR-CNN - Multi-Object Tracking and Segmentation](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FTrackR-CNN) [cvpr19] [tensorflow] [[project]](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fmots) \n* [Eliminating Exposure Bias and Metric Mismatch in Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002Fmaksay\u002Fseq-train) [cvpr19] [tensorflow]    \n* [Robust Multi-Modality Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002FZwwWayne\u002FmmMOT) [iccv19] [pytorch]    \n* [Towards Real-Time Multi-Object Tracking \u002F Joint Detection and Embedding](https:\u002F\u002Fgithub.com\u002FZhongdao\u002FTowards-Realtime-MOT) [ax1909] [pytorch] [[CMU]](https:\u002F\u002Fgithub.com\u002FJunweiLiang\u002FObject_Detection_Tracking)\n* [Deep Affinity Network for Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002FshijieS\u002FSST) [tpami19] [pytorch]    \n* [Tracking without bells and whistles](https:\u002F\u002Fgithub.com\u002Fphil-bergmann\u002Ftracking_wo_bnw) [iccv19] [pytorch]    \n* [Lifted Disjoint Paths with Application in Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002FAndreaHor\u002FLifT_Solver) [icml20] [matlab] [mot15#1,mot16 #3,mot17#2]   \n* [Learning a Neural Solver for Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002Fdvl-tum\u002Fmot_neural_solver) [cvpr20] [pytorch] [mot15#2]   \n* [Tracking Objects as Points](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterTrack) [ax2004] [pytorch]\n* [Quasi-Dense Similarity Learning for Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002FSysCV\u002Fqdtrack) [ax2006] [pytorch]\n* [DEFT: Detection Embeddings for Tracking](https:\u002F\u002Fgithub.com\u002FMedChaabane\u002FDEFT) [ax2102] [pytorch]\n* [How To Train Your Deep Multi-Object Tracker](https:\u002F\u002Fgithub.com\u002FyihongXU\u002FdeepMOT) [ax1906\u002Fcvpr20] [pytorch] [[traktor\u002Fgitlab]](https:\u002F\u002Fgitlab.inria.fr\u002Fyixu\u002Fdeepmot)\n* [Track To Detect and Segment: An Online Multi-Object Tracker ](https:\u002F\u002Fgithub.com\u002FJialianW\u002FTraDeS) [cvpr21] [pytorch] [[project]](https:\u002F\u002Fjialianwu.com\u002Fprojects\u002FTraDeS.html)\n* [MOTR: End-to-End Multiple-Object Tracking with Transformer](https:\u002F\u002Fgithub.com\u002Fmegvii-model\u002FMOTR) [ax2202] [pytorch]\n\n\n\u003Ca id=\"baselin_e__1\">\u003C\u002Fa>\n### Baseline\n* [Simple Online and Realtime Tracking](https:\u002F\u002Fgithub.com\u002Fabewley\u002Fsort) [icip 2016] [python]\n* [Deep SORT : Simple Online Realtime Tracking with a Deep Association Metric](https:\u002F\u002Fgithub.com\u002Fnwojke\u002Fdeep_sort) [icip17] [python]\n* [High-Speed Tracking-by-Detection Without Using Image Information](https:\u002F\u002Fgithub.com\u002Fbochinski\u002Fiou-tracker) [avss17] [python]  \n* [A simple baseline for one-shot multi-object tracking](https:\u002F\u002Fgithub.com\u002Fifzhang\u002FFairMOT) [ax2004] [pytorch] [winner of mot15,16,17,20]\n\n\u003Ca id=\"siamese__1\">\u003C\u002Fa>\n### Siamese\n* [SiamMOT: Siamese Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002Famazon-research\u002Fsiam-mot) [ax2105] [pytorch]\n    \n\u003Ca id=\"unsupervise_d_\">\u003C\u002Fa>\n### Unsupervised\n* [Tracking by Animation: Unsupervised Learning of Multi-Object Attentive Trackers](https:\u002F\u002Fgithub.com\u002Fzhen-he\u002Ftracking-by-animation) [cvpr19] [python\u002Fc++\u002Fpytorch]\n    \n\u003Ca id=\"re_id_\">\u003C\u002Fa>\n### Re-ID\n* [Torchreid: Deep learning person re-identification in PyTorch](https:\u002F\u002Fgithub.com\u002FKaiyangZhou\u002Fdeep-person-reid) [ax1910] [pytorch]\n* [SMOT: Single-Shot Multi Object Tracking](https:\u002F\u002Fgithub.com\u002Fdmlc\u002Fgluon-cv\u002Ftree\u002Fmaster\u002Fgluoncv\u002Fmodel_zoo\u002Fsmot) [ax2010] [pytorch] [gluon-cv]\n* [FairMOT: On the Fairness of Detection and Re-Identification in Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002Fifzhang\u002FFairMOT) [ax2004] [pytorch] [[microsoft]](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FFairMOT) [[BDD100K]](https:\u002F\u002Fgithub.com\u002Fdingwoai\u002FFairMOT-BDD100K) [[face tracking]](https:\u002F\u002Fgithub.com\u002Fzengwb-lx\u002FFace-Tracking-usingFairMOT)\n* [Rethinking the competition between detection and ReID in Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002FJudasDie\u002FSOTS) [ax2010] [pytorch] \n\n\u003Ca id=\"framework_s__1\">\u003C\u002Fa>\n#### Frameworks\n* [PyTorch open-source toolbox for unsupervised or domain adaptive object re-ID](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002FOpenUnReID) [pytorch] \n\n\n\u003Ca id=\"graph_nn_\">\u003C\u002Fa>\n### Graph NN\n* [Joint Object Detection and Multi-Object Tracking with Graph Neural Networks](https:\u002F\u002Fgithub.com\u002Fyongxinw\u002FGSDT) [ax2006\u002F icra21] [pytorch]\n\n\u003Ca id=\"microscopy___cell_tracking__1\">\u003C\u002Fa>\n### Microscopy \u002F cell tracking\n* [Baxter Algorithms \u002F Viterbi Tracking](https:\u002F\u002Fgithub.com\u002Fklasma\u002FBaxterAlgorithms) [tmi14] [matlab]\n* [Deepcell: Accurate cell tracking and lineage construction in live-cell imaging experiments with deep learning](https:\u002F\u002Fgithub.com\u002Fvanvalenlab\u002Fdeepcell-tracking) [biorxiv1910] [tensorflow]\n\n\u003Ca id=\"3_d_\">\u003C\u002Fa>\n### 3D\n* [3D Multi-Object Tracking: A Baseline and New Evaluation Metrics ](https:\u002F\u002Fgithub.com\u002Fxinshuoweng\u002FAB3DMOT) [iros20\u002Feccvw20] [pytorch]\n* [GNN3DMOT: Graph Neural Network for 3D Multi-Object Tracking with Multi-Feature Learning ](https:\u002F\u002Fgithub.com\u002Fxinshuoweng\u002FGNN3DMOT) [iros20\u002Feccvw20] [pytorch]\n\n\u003Ca id=\"metrics__1\">\u003C\u002Fa>\n### Metrics\n* [HOTA: A Higher Order Metric for Evaluating Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002FJonathonLuiten\u002FHOTA-metrics) [cvpr20] [python]\n\n\u003Ca id=\"single_object_tracking__2\">\u003C\u002Fa>\n## Single Object Tracking\n* [A collection of common tracking algorithms (2003-2012)](https:\u002F\u002Fgithub.com\u002Fzenhacker\u002FTrackingAlgoCollection) [c++\u002Fmatlab]\n* [SenseTime Research platform for single object tracking, implementing algorithms like SiamRPN and SiamMask](https:\u002F\u002Fgithub.com\u002FSTVIR\u002Fpysot\u002F) [pytorch]\n* [In Defense of Color-based Model-free Tracking](https:\u002F\u002Fgithub.com\u002Fjbhuang0604\u002FCF2) [cvpr15] [c++]\n* [Hierarchical Convolutional Features for Visual Tracking](https:\u002F\u002Fgithub.com\u002Fjbhuang0604\u002FCF2) [iccv15] [matlab]\n* [Visual Tracking with Fully Convolutional Networks](https:\u002F\u002Fgithub.com\u002Fscott89\u002FFCNT) [iccv15] [matlab]\n* [Hierarchical Convolutional Features for Visual Tracking](https:\u002F\u002Fgithub.com\u002Fjbhuang0604\u002FCF2) [iccv15] [matlab] \n* [DeepTracking: Seeing Beyond Seeing Using Recurrent Neural Networks](https:\u002F\u002Fgithub.com\u002Fpondruska\u002FDeepTracking) [aaai16] [torch 7]\n* Learning Multi-Domain Convolutional Neural Networks for Visual Tracking [cvpr16] [vot2015 winner] [[matlab\u002Fmatconvnet]](https:\u002F\u002Fgithub.com\u002FHyeonseobNam\u002FMDNet) [[pytorch]](https:\u002F\u002Fgithub.com\u002FHyeonseobNam\u002Fpy-MDNet)\n* [Beyond Correlation Filters: Learning Continuous Convolution Operators for Visual Tracking](https:\u002F\u002Fgithub.com\u002Fmartin-danelljan\u002FContinuous-ConvOp) [eccv 2016] [matlab]\n* [Fully-Convolutional Siamese Networks for Object Tracking](https:\u002F\u002Fgithub.com\u002Fbertinetto\u002Fsiamese-fc) [eccvw 2016] [matlab\u002Fmatconvnet] [[project]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~luca\u002Fsiamese-fc.html) [[pytorch]](https:\u002F\u002Fgithub.com\u002Fhuanglianghua\u002Fsiamfc-pytorch) [[pytorch (only training)]](https:\u002F\u002Fgithub.com\u002Frafellerc\u002FPytorch-SiamFC)\n* [DCFNet: Discriminant Correlation Filters Network for Visual Tracking](https:\u002F\u002Farxiv.org\u002Fabs\u002F1704.04057) [ax1704] [[matlab\u002Fmatconvnet]](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FDCFNet\u002F) [[pytorch]](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FDCFNet_pytorch\u002F)\n* [End-to-end representation learning for Correlation Filter based tracking](https:\u002F\u002Farxiv.org\u002Fabs\u002F1704.06036)\n[cvpr17]\n[[matlab\u002Fmatconvnet]](https:\u002F\u002Fgithub.com\u002Fbertinetto\u002Fcfnet) [[tensorflow\u002Finference_only]](https:\u002F\u002Fgithub.com\u002Ftorrvision\u002Fsiamfc-tf) [[project]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~luca\u002Fsiamese-fc.html)\n* [Dual Deep Network for Visual Tracking](https:\u002F\u002Fgithub.com\u002Fchizhizhen\u002FDNT) [tip1704] [caffe]\n* [SiameseX: A simplified PyTorch implementation of Siamese networks for tracking: SiamFC, SiamRPN, SiamRPN++, SiamVGG, SiamDW, SiamRPN-VGG](https:\u002F\u002Fgithub.com\u002Fzllrunning\u002FSiameseX.PyTorch) [pytorch]\n* [RATM: Recurrent Attentive Tracking Model](https:\u002F\u002Fgithub.com\u002Fsaebrahimi\u002FRATM) [cvprw17] [python]\n* [ROLO : Spatially Supervised Recurrent Convolutional Neural Networks for Visual Object Tracking](https:\u002F\u002Fgithub.com\u002FGuanghan\u002FROLO) [iscas 2017] [tensorfow]\n* [ECO: Efficient Convolution Operators for Tracking](https:\u002F\u002Farxiv.org\u002Fabs\u002F1611.09224)\n[cvpr17]\n[[matlab]](https:\u002F\u002Fgithub.com\u002Fmartin-danelljan\u002FECO)\n[[python\u002Fcuda]](https:\u002F\u002Fgithub.com\u002FStrangerZhang\u002FpyECO)\n[[pytorch]](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking)\n* [Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning](https:\u002F\u002Fgithub.com\u002Fildoonet\u002Ftf-adnet-tracking) [cvpr17] [tensorflow]\n* [Detect to Track and Track to Detect](https:\u002F\u002Fgithub.com\u002Ffeichtenhofer\u002FDetect-Track) [iccv17] [matlab]\n* [Meta-Tracker: Fast and Robust Online Adaptation for Visual Object Trackers](https:\u002F\u002Fgithub.com\u002Fsilverbottlep\u002Fmeta_trackers) [eccv18] [pytorch]\n* [Learning Spatial-Temporal Regularized Correlation Filters for Visual Tracking](https:\u002F\u002Fgithub.com\u002Flifeng9472\u002FSTRCF) [cvpr18] [matlab]\n* High Performance Visual Tracking with Siamese Region Proposal Network [cvpr18] [[pytorch\u002F195]](https:\u002F\u002Fgithub.com\u002Fzkisthebest\u002FSiamese-RPN) [[pytorch\u002F313]](https:\u002F\u002Fgithub.com\u002Fsongdejia\u002FSiamese-RPN-pytorch)  [[pytorch\u002Fno_train\u002F104]](https:\u002F\u002Fgithub.com\u002Fhuanglianghua\u002Fsiamrpn-pytorch) [[pytorch\u002F177]](https:\u002F\u002Fgithub.com\u002FHelloRicky123\u002FSiamese-RPN) \n* [Distractor-aware Siamese Networks for Visual Object Tracking](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FDaSiamRPN) [eccv18] [vot18 winner] [pytorch]\n* [VITAL: VIsual Tracking via Adversarial Learning](https:\u002F\u002Fgithub.com\u002Fybsong00\u002FVital_release) [cvpr18] [matlab] [[pytorch]](https:\u002F\u002Fgithub.com\u002Fabnerwang\u002Fpy-Vital) [[project]](https:\u002F\u002Fybsong00.github.io\u002Fcvpr18_tracking\u002Findex.html)\n* [Fast Online Object Tracking and Segmentation: A Unifying Approach (SiamMask)](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FSiamMask) [cvpr19] [pytorch] [[project]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~qwang\u002FSiamMask\u002F)\n* [PyTracking: A general python framework for training and running visual object trackers, based on PyTorch](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking) [ECO\u002FATOM\u002FDiMP\u002FPrDiMP] [cvpr17\u002Fcvpr19\u002Ficcv19\u002Fcvpr20] [pytorch] \n* [Unsupervised Deep Tracking](https:\u002F\u002Fgithub.com\u002F594422814\u002FUDT) [cvpr19] [matlab\u002Fmatconvnet] [[pytorch]](https:\u002F\u002Fgithub.com\u002F594422814\u002FUDT_pytorch)\n* [Deeper and Wider Siamese Networks for Real-Time Visual Tracking](https:\u002F\u002Fgithub.com\u002Fresearchmm\u002FSiamDW) [cvpr19] [pytorch]\n* [GradNet: Gradient-Guided Network for Visual Object Tracking](https:\u002F\u002Fgithub.com\u002FLPXTT\u002FGradNet-Tensorflow) [iccv19] [tensorflow]\n* [`Skimming-Perusal' Tracking: A Framework for Real-Time and Robust Long-term Tracking](https:\u002F\u002Fgithub.com\u002Fiiau-tracker\u002FSPLT) [iccv19] [tensorflow]\n* [Learning Aberrance Repressed Correlation Filters for Real-Time UAV Tracking](https:\u002F\u002Fgithub.com\u002Fvision4robotics\u002FARCF-tracker) [iccv19] [matlab]\n* [Learning the Model Update for Siamese Trackers](https:\u002F\u002Fgithub.com\u002Fzhanglichao\u002Fupdatenet) [iccv19] [pytorch]\n* [SPM-Tracker: Series-Parallel Matching for Real-Time Visual Object Tracking](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FSPM-Tracker) [cvpr19] [pytorch] [inference-only]\n* [Joint Group Feature Selection and Discriminative Filter Learning for Robust Visual Object Tracking](https:\u002F\u002Fgithub.com\u002FXU-TIANYANG\u002FGFS-DCF) [iccv19] [matlab]\n* [Siam R-CNN: Visual Tracking by Re-Detection](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FSiamR-CNN) [cvpr20] [tensorflow]\n* [D3S - Discriminative Single Shot Segmentation Tracker](https:\u002F\u002Fgithub.com\u002Falanlukezic\u002Fd3s) [cvpr20] [pytorch\u002Fpytracking]\n* [Discriminative and Robust Online Learning for Siamese Visual Tracking](https:\u002F\u002Fgithub.com\u002Fshallowtoil\u002FDROL) [aaai20] [pytorch\u002Fpysot]\n* [Siamese Box Adaptive Network for Visual Tracking](https:\u002F\u002Fgithub.com\u002Fhqucv\u002Fsiamban) [cvpr20] [pytorch\u002Fpysot]\n* [Ocean: Object-aware Anchor-free Tracking](https:\u002F\u002Fgithub.com\u002FJudasDie\u002FSOTS) [ax2010] [pytorch] \n\n\u003Ca id=\"gui_application___large_scale_tracking___animal_s_\">\u003C\u002Fa>\n### GUI Application \u002F Large Scale Tracking \u002F Animals\n* [BioTracker An Open-Source Computer Vision Framework for Visual Animal Tracking](https:\u002F\u002Fgithub.com\u002FBioroboticsLab\u002Fbiotracker_core)[opencv\u002Fc++]\n* [Tracktor: Image‐based automated tracking of animal movement and behaviour](https:\u002F\u002Fgithub.com\u002Fvivekhsridhar\u002Ftracktor)[opencv\u002Fc++]\n* [MARGO (Massively Automated Real-time GUI for Object-tracking), a platform for high-throughput ethology](https:\u002F\u002Fgithub.com\u002Fde-Bivort-Lab\u002Fmargo)[matlab]\n* [idtracker.ai: Tracking all individuals in large collectives of unmarked animals](https:\u002F\u002Fgitlab.com\u002Fpolavieja_lab\u002Fidtrackerai)\n[tensorflow]\n[[project]](https:\u002F\u002Fidtracker.ai\u002F)\n    \n\u003Ca id=\"video_detectio_n__2\">\u003C\u002Fa>\n## Video Detection\n* [Flow-Guided Feature Aggregation for Video Object Detection](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FFlow-Guided-Feature-Aggregation)\n[nips16 \u002F iccv17]\n[mxnet]\n* [T-CNN: Tubelets with Convolution Neural Networks](https:\u002F\u002Fgithub.com\u002Fmyfavouritekk\u002FT-CNN) [cvpr16] [python]  \n* [TPN: Tubelet Proposal Network](https:\u002F\u002Fgithub.com\u002Fmyfavouritekk\u002FTPN) [cvpr17] [python]\n* [Deep Feature Flow for Video Recognition](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FDeep-Feature-Flow) [cvpr17] [mxnet]\n* [Mobile Video Object Detection with Temporally-Aware Feature Maps](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Flstm_object_detection) [cvpr18] [Google] [tensorflow]  \n\n\u003Ca id=\"action_detectio_n_\">\u003C\u002Fa>\n### Action Detection\n\u003Ca id=\"framework_s__2\">\u003C\u002Fa>\n#### Frameworks\n+ [OpenMMLab's Next Generation Video Understanding Toolbox and Benchmark](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmaction2) [pytorch]\n\n\u003Ca id=\"static_detection_and_matching_\">\u003C\u002Fa>\n## Static Detection and Matching\n\u003Ca id=\"framework_s__3\">\u003C\u002Fa>\n### Frameworks\n+ [Tensorflow object detection API](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fobject_detection) [tensorflow]\n+ [Detectron2](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetectron2) [pytorch]\n+ [Detectron](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDetectron) [pytorch]\n+ [Open MMLab Detection Toolbox with PyTorch](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmdetection) [pytorch]\n+ [SimpleDet](https:\u002F\u002Fgithub.com\u002Ftusen-ai\u002Fsimpledet) [mxnet]\n\n\u003Ca id=\"region_proposal__1\">\u003C\u002Fa>\n### Region Proposal   \n+ [MCG : Multiscale Combinatorial Grouping - Object Proposals and Segmentation](https:\u002F\u002Fgithub.com\u002Fjponttuset\u002Fmcg)  [(project)](https:\u002F\u002Fwww2.eecs.berkeley.edu\u002FResearch\u002FProjects\u002FCS\u002Fvision\u002Fgrouping\u002Fmcg\u002F) [tpami16\u002Fcvpr14] [python]\n+ [COB : Convolutional Oriented Boundaries](https:\u002F\u002Fgithub.com\u002Fkmaninis\u002FCOB)  [(project)](http:\u002F\u002Fwww.vision.ee.ethz.ch\u002F~cvlsegmentation\u002Fcob\u002F) [tpami18\u002Feccv16] [matlab\u002Fcaffe]\n\n\u003Ca id=\"fpn_\">\u003C\u002Fa>\n### FPN\n* [Feature Pyramid Networks for Object Detection](https:\u002F\u002Fgithub.com\u002Funsky\u002FFPN) [caffe\u002Fpython]  \n\n\u003Ca id=\"rcn_n__1\">\u003C\u002Fa>\n### RCNN\n* [RFCN (author)](https:\u002F\u002Fgithub.com\u002Fdaijifeng001\u002Fr-fcn) [caffe\u002Fmatlab]\n* [RFCN-tensorflow](https:\u002F\u002Fgithub.com\u002Fxdever\u002FRFCN-tensorflow) [tensorflow]\n* [PVANet: Lightweight Deep Neural Networks for Real-time Object Detection](https:\u002F\u002Fgithub.com\u002Fsanghoon\u002Fpva-faster-rcnn) [intel] [emdnn16(nips16)]\n* Mask R-CNN [[tensorflow]](https:\u002F\u002Fgithub.com\u002FCharlesShang\u002FFastMaskRCNN) [[keras]](https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)\n* [Light-head R-CNN](https:\u002F\u002Fgithub.com\u002Fzengarden\u002Flight_head_rcnn) [cvpr18] [tensorflow]    \n* [Evolving Boxes for Fast Vehicle Detection](https:\u002F\u002Fgithub.com\u002FWilly0919\u002FEvolving_Boxes) [icme18] [caffe\u002Fpython]\n* [Cascade R-CNN (cvpr18)](http:\u002F\u002Fwww.svcl.ucsd.edu\u002Fpublications\u002Fconference\u002F2018\u002Fcvpr\u002Fcascade-rcnn.pdf) [[detectron]](https:\u002F\u002Fgithub.com\u002Fzhaoweicai\u002FDetectron-Cascade-RCNN) [[caffe]](https:\u002F\u002Fgithub.com\u002Fzhaoweicai\u002Fcascade-rcnn)  \n* [A MultiPath Network for Object Detection](https:\u002F\u002Farxiv.org\u002Fabs\u002F1604.02135) [[torch]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fmultipathnet) [bmvc16] [facebook]\n* [SNIPER: Efficient Multi-Scale Training\u002FAn Analysis of Scale Invariance in Object Detection-SNIP](https:\u002F\u002Fgithub.com\u002Fmahyarnajibi\u002FSNIPER) [nips18\u002Fcvpr18] [mxnet]\n\n\u003Ca id=\"ssd__1\">\u003C\u002Fa>\n### SSD\n* [SSD-Tensorflow](https:\u002F\u002Fgithub.com\u002Fljanyst\u002Fssd-tensorflow) [tensorflow]\n* [SSD-Tensorflow (tf.estimator)](https:\u002F\u002Fgithub.com\u002FHiKapok\u002FSSD.TensorFlow) [tensorflow]\n* [SSD-Tensorflow (tf.slim)](https:\u002F\u002Fgithub.com\u002Fbalancap\u002FSSD-Tensorflow) [tensorflow]\n* [SSD-Keras](https:\u002F\u002Fgithub.com\u002Frykov8\u002Fssd_keras) [keras]\n* [SSD-Pytorch](https:\u002F\u002Fgithub.com\u002Famdegroot\u002Fssd.pytorch) [pytorch]\n* [Enhanced SSD with Feature Fusion and Visual Reasoning](https:\u002F\u002Fgithub.com\u002FCVlengjiaxu\u002FEnhanced-SSD-with-Feature-Fusion-and-Visual-Reasoning) [nca18] [tensorflow]\n* [RefineDet - Single-Shot Refinement Neural Network for Object Detection](https:\u002F\u002Fgithub.com\u002Fsfzhang15\u002FRefineDet) [cvpr18] [caffe]\n\n\u003Ca id=\"retinanet__1\">\u003C\u002Fa>\n### RetinaNet\n* [9.277.41](https:\u002F\u002Fgithub.com\u002Fc0nn3r\u002FRetinaNet) [pytorch]\n* [31.857.212](https:\u002F\u002Fgithub.com\u002Fkuangliu\u002Fpytorch-retinanet) [pytorch]\n* [25.274.84](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fretinanet-examples) [pytorch] [nvidia]\n* [22.869.302](https:\u002F\u002Fgithub.com\u002Fyhenon\u002Fpytorch-retinanet) [pytorch]\n\n\u003Ca id=\"yol_o__1\">\u003C\u002Fa>\n### YOLO  \n+ [Darknet: Convolutional Neural Networks](https:\u002F\u002Fgithub.com\u002Fpjreddie\u002Fdarknet) [c\u002Fpython]\n+ [YOLO9000: Better, Faster, Stronger - Real-Time Object Detection. 9000 classes!](https:\u002F\u002Fgithub.com\u002Fphilipperemy\u002Fyolo-9000)  [c\u002Fpython]\n+ [Darkflow](https:\u002F\u002Fgithub.com\u002Fthtrieu\u002Fdarkflow) [tensorflow]\n+ [Pytorch Yolov2 ](https:\u002F\u002Fgithub.com\u002Fmarvis\u002Fpytorch-yolo2) [pytorch]\n+ [Yolo-v3 and Yolo-v2 for Windows and Linux](https:\u002F\u002Fgithub.com\u002FAlexeyAB\u002Fdarknet) [c\u002Fpython]\n+ [YOLOv3 in PyTorch](https:\u002F\u002Fgithub.com\u002Fultralytics\u002Fyolov3) [pytorch]\n+ [pytorch-yolo-v3 ](https:\u002F\u002Fgithub.com\u002Fayooshkathuria\u002Fpytorch-yolo-v3) [pytorch] [no training] [[tutorial]](https:\u002F\u002Fblog.paperspace.com\u002Fhow-to-implement-a-yolo-object-detector-in-pytorch\u002F)\n+ [YOLOv3_TensorFlow](https:\u002F\u002Fgithub.com\u002Fwizyoung\u002FYOLOv3_TensorFlow) [tensorflow]\n+ [tensorflow-yolo-v3](https:\u002F\u002Fgithub.com\u002Fmystic123\u002Ftensorflow-yolo-v3) [tensorflow slim]\n+ [tensorflow-yolov3](https:\u002F\u002Fgithub.com\u002FYunYang1994\u002Ftensorflow-yolov3) [tensorflow slim]\n+ [keras-yolov3](https:\u002F\u002Fgithub.com\u002Fqqwweee\u002Fkeras-yolo3) [keras]  \n+ YOLOv4 [[darknet - c\u002Fpython]](https:\u002F\u002Fgithub.com\u002FAlexeyAB\u002Fdarknet) [[tensorflow]](https:\u002F\u002Fgithub.com\u002Fhunglc007\u002Ftensorflow-yolov4-tflite) [[pytorch\u002F711]](https:\u002F\u002Fgithub.com\u002FWongKinYiu\u002FPyTorch_YOLOv4) [[pytorch\u002FONNX\u002FTensorRT\u002F1.9k]](https:\u002F\u002Fgithub.com\u002FTianxiaomo\u002Fpytorch-YOLOv4) [[pytorch 3D]](https:\u002F\u002Fgithub.com\u002Fmaudzung\u002FComplex-YOLOv4-Pytorch)\n+ [YOLOv5](https:\u002F\u002Fgithub.com\u002Fultralytics\u002Fyolov5) [pytorch] \n+ [YOLOX](https:\u002F\u002Fgithub.com\u002FMegvii-BaseDetection\u002FYOLOX) [pytorch] [MegEngine](https:\u002F\u002Fgithub.com\u002FMegEngine\u002FYOLOX) [ax2107]\n\n\u003Ca id=\"anchor_free__1\">\u003C\u002Fa>\n### Anchor Free\n* [FoveaBox: Beyond Anchor-based Object Detector](https:\u002F\u002Fgithub.com\u002Ftaokong\u002FFoveaBox) [ax1904] [pytorch\u002Fmmdetection]\n* [Cornernet: Detecting objects as paired keypoints](https:\u002F\u002Fgithub.com\u002Fprinceton-vl\u002FCornerNet) [ax1903\u002Feccv18] [pytorch]\n* [FCOS: Fully Convolutional One-Stage Object Detection](https:\u002F\u002Fgithub.com\u002Ftianzhi0549\u002FFCOS) [iccv19] [pytorch] [[VoVNet]](https:\u002F\u002Fgithub.com\u002Fvov-net\u002FVoVNet-FCOS) [[HRNet]](https:\u002F\u002Fgithub.com\u002FHRNet\u002FHRNet-FCOS) [[NAS]](https:\u002F\u002Fgithub.com\u002FLausannen\u002FNAS-FCOS) [[FCOS_PLUS]](https:\u002F\u002Fgithub.com\u002Fyqyao\u002FFCOS_PLUS)\n* [Feature Selective Anchor-Free Module for Single-Shot Object Detection](https:\u002F\u002Fgithub.com\u002Fhdjang\u002FFeature-Selective-Anchor-Free-Module-for-Single-Shot-Object-Detection) [cvpr19] [pytorch]\n* [CenterNet: Objects as Points](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterNet) [ax1904] [pytorch]\n* [Bottom-up Object Detection by Grouping Extreme and Center Points,](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FExtremeNet) [cvpr19]  [pytorch]\n* [RepPoints Point Set Representation for Object Detection](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FRepPoints) [iccv19]  [pytorch] [microsoft]\n* [DE⫶TR: End-to-End Object Detection with Transformers](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetr) [ax200528]  [pytorch] [facebook]\n* [Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection](https:\u002F\u002Fgithub.com\u002Fsfzhang15\u002FATSS) [cvpr20]  [pytorch]\n\n\u003Ca id=\"mis_c__2\">\u003C\u002Fa>\n### Misc\n* [Relation Networks for Object Detection](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FRelation-Networks-for-Object-Detection) [cvpr18] [mxnet]\n* [DeNet: Scalable Real-time Object Detection with Directed Sparse Sampling](https:\u002F\u002Fgithub.com\u002Flachlants\u002Fdenet) [iccv17(poster)] [theano]\n* [Multi-scale Location-aware Kernel Representation for Object Detection](https:\u002F\u002Fgithub.com\u002FHwang64\u002FMLKP) [cvpr18]  [caffe\u002Fpython]\n\n\u003Ca id=\"matchin_g_\">\u003C\u002Fa>\n### Matching  \n+ [Matchnet](https:\u002F\u002Fgithub.com\u002Fhanxf\u002Fmatchnet)\n+ [Stereo Matching by Training a Convolutional Neural Network to Compare Image Patches](https:\u002F\u002Fgithub.com\u002Fjzbontar\u002Fmc-cnn)\n\n\u003Ca id=\"boundary_detectio_n__1\">\u003C\u002Fa>\n### Boundary Detection  \n+ [Holistically-Nested Edge Detection (HED) (iccv15)](https:\u002F\u002Fgithub.com\u002Fs9xie\u002Fhed) [caffe]       \n+ [Edge-Detection-using-Deep-Learning (HED)](https:\u002F\u002Fgithub.com\u002FAkuanchang\u002FEdge-Detection-using-Deep-Learning) [tensorflow]\n+ [Holistically-Nested Edge Detection (HED) in OpenCV](https:\u002F\u002Fgithub.com\u002Fopencv\u002Fopencv\u002Fblob\u002Fmaster\u002Fsamples\u002Fdnn\u002Fedge_detection.py) [python\u002Fc++]       \n+ [Crisp Boundary Detection Using Pointwise Mutual Information (eccv14)](https:\u002F\u002Fgithub.com\u002Fphillipi\u002Fcrisp-boundaries) [matlab]\n+ [Dense Extreme Inception Network: Towards a Robust CNN Model for Edge Detection](https:\u002F\u002Fgithub.com\u002Fphillipi\u002Fcrisp-boundaries) [wacv20] [tensorflow](https:\u002F\u002Fgithub.com\u002Fxavysp\u002FDexiNed\u002Ftree\u002Fmaster\u002Flegacy) [pytorch](https:\u002F\u002Fgithub.com\u002Fxavysp\u002FDexiNed)\n\n\u003Ca id=\"text_detectio_n_\">\u003C\u002Fa>\n### Text Detection  \n+ [Real-time Scene Text Detection with Differentiable Binarization](https:\u002F\u002Fgithub.com\u002FMhLiao\u002FDB) [pytorch] [aaai20] \n\n\u003Ca id=\"framework_s__4\">\u003C\u002Fa>\n#### Frameworks\n+ [ OpenMMLab Text Detection, Recognition and Understanding Toolbox ](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmocr) [pytorch]\n\n\u003Ca id=\"3d_detectio_n_\">\u003C\u002Fa>\n### 3D Detection  \n\u003Ca id=\"framework_s__5\">\u003C\u002Fa>\n#### Frameworks\n+ [OpenMMLab's next-generation platform for general 3D object detection](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmdetection3d) [pytorch]\n+ [OpenPCDet Toolbox for LiDAR-based 3D Object Detection](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002FOpenPCDet) [pytorch]\n\n\u003Ca id=\"optical_flow__1\">\u003C\u002Fa>\n## Optical Flow\n* [FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks (cvpr17)](https:\u002F\u002Farxiv.org\u002Fabs\u002F1612.01925) [[caffe]](https:\u002F\u002Fgithub.com\u002Flmb-freiburg\u002Fflownet2) [[pytorch\u002Fnvidia]](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fflownet2-pytorch)\n* [SPyNet: Spatial Pyramid Network for Optical Flow (cvpr17)](https:\u002F\u002Farxiv.org\u002Fabs\u002F1702.02295) [[lua]](https:\u002F\u002Fgithub.com\u002Fanuragranj\u002Fspynet) [[pytorch]](https:\u002F\u002Fgithub.com\u002Fsniklaus\u002Fpytorch-spynet)\n* [Guided Optical Flow Learning (cvprw17)](https:\u002F\u002Farxiv.org\u002Fabs\u002F1702.02295) [[caffe]](https:\u002F\u002Fgithub.com\u002Fbryanyzhu\u002FGuidedNet) [[tensorflow]](https:\u002F\u002Fgithub.com\u002Fbryanyzhu\u002FdeepOF)\n* [Fast Optical Flow using Dense Inverse Search (DIS)](https:\u002F\u002Fgithub.com\u002Ftikroeger\u002FOF_DIS) [eccv16] [C++]\n* [A Filter Formulation for Computing Real Time Optical Flow](https:\u002F\u002Fgithub.com\u002Fjadarve\u002Foptical-flow-filter) [ral16] [c++\u002Fcuda - matlab,python wrappers]\n* [PatchBatch - a Batch Augmented Loss for Optical Flow](https:\u002F\u002Fgithub.com\u002FDediGadot\u002FPatchBatch) [cvpr16] [python\u002Ftheano]\n* [Piecewise Rigid Scene Flow](https:\u002F\u002Fgithub.com\u002Fvogechri\u002FPRSM) [iccv13\u002Feccv14\u002Fijcv15] [c++\u002Fmatlab]\n* [DeepFlow v2](https:\u002F\u002Farxiv.org\u002Fabs\u002F1611.00850) [iccv13] [[c++\u002Fpython\u002Fmatlab]](https:\u002F\u002Fgithub.com\u002Fzimenglan-sysu-512\u002Fdeep-flow), [[project]](http:\u002F\u002Flear.inrialpes.fr\u002Fsrc\u002Fdeepflow\u002F)\n* [An Evaluation of Data Costs for Optical Flow](https:\u002F\u002Fgithub.com\u002Fvogechri\u002FDataFlow) [gcpr13] [matlab]\n\n\u003Ca id=\"framework_s__6\">\u003C\u002Fa>\n### Frameworks\n+ [OpenMMLab optical flow toolbox and benchmark ](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmflow) [pytorch]\n\n\n\u003Ca id=\"instance_segmentation_\">\u003C\u002Fa>\n## Instance Segmentation\n* [Fully Convolutional Instance-aware Semantic Segmentation](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FFCIS) [cvpr17] [coco16 winner] [mxnet]\n* [Instance-aware Semantic Segmentation via Multi-task Network Cascades](https:\u002F\u002Fgithub.com\u002Fdaijifeng001\u002FMNC) [cvpr16] [caffe] [coco15 winner]    \n* [DeepMask\u002FSharpMask](https:\u002F\u002Farxiv.org\u002Fabs\u002F1603.08695) [nips15\u002Feccv16] [facebook] [[torch]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdeepmask) [[tensorflow]](https:\u002F\u002Fgithub.com\u002Faby2s\u002Fsharpmask)  [[pytorch\u002Fdeepmask]](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002Fdeepmask-pytorch\u002F) \n* [Simultaneous Detection and Segmentation](https:\u002F\u002Fgithub.com\u002Fbharath272\u002Fsds_eccv2014) [eccv14] [matlab] [[project]](https:\u002F\u002Fwww2.eecs.berkeley.edu\u002FResearch\u002FProjects\u002FCS\u002Fvision\u002Fshape\u002Fsds\u002F)    \n* [PANet](https:\u002F\u002Fgithub.com\u002FShuLiu1993\u002FPANet) [cvpr18] [pytorch]\n* [RetinaMask](https:\u002F\u002Fgithub.com\u002Fchengyangfu\u002Fretinamask) [arxviv1901] [pytorch]\n* [Mask Scoring R-CNN](https:\u002F\u002Fgithub.com\u002Fzjhuang22\u002Fmaskscoring_rcnn) [cvpr19] [pytorch]\n* [DeepMAC](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Fblob\u002Fmaster\u002Fresearch\u002Fobject_detection\u002Fg3doc\u002Fdeepmac.md) [ax2104] [tensorflow]\n* [Swin Transformer](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FSwin-Transformer) [iccv21] [pytorch] [microsoft]\n\n\u003Ca id=\"framework_s__7\">\u003C\u002Fa>\n### Frameworks\n* [Fast, modular reference implementation of Instance Segmentation and Object Detection algorithms in PyTorch](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fmaskrcnn-benchmark) [pytorch] [facebook]\n* [PaddleDetection, Object detection and instance segmentation toolkit based on PaddlePaddle.](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleDetection) [2019]\n\n\u003Ca id=\"semantic_segmentation_\">\u003C\u002Fa>\n## Semantic Segmentation\n* [Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation](https:\u002F\u002Fgithub.com\u002Fswamiviv\u002FLSD-seg) [cvpr18] [spotlight] [pytorch]\n* [Few-shot Segmentation Propagation with Guided Networks](https:\u002F\u002Fgithub.com\u002Fshelhamer\u002Frevolver) [ax1806] [pytorch] [incomplete]\n* [Pytorch-segmentation-toolbox](https:\u002F\u002Fgithub.com\u002Fspeedinghzl\u002Fpytorch-segmentation-toolbox) [DeeplabV3 and PSPNet] [pytorch]\n* [DeepLab](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fdeeplab) [tensorflow]\n* [Auto-DeepLab](https:\u002F\u002Fgithub.com\u002FMenghaoGuo\u002FAutoDeeplab) [pytorch]\n* [DeepLab v3+](https:\u002F\u002Fgithub.com\u002Fjfzhang95\u002Fpytorch-deeplab-xception) [pytorch]\n* [Deep Extreme Cut (DEXTR): From Extreme Points to Object Segmentation](https:\u002F\u002Fgithub.com\u002Fscaelles\u002FDEXTR-PyTorch)[cvpr18][[project]](https:\u002F\u002Fcvlsegmentation.github.io\u002Fdextr\u002F) [pytorch]\n* [FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation](https:\u002F\u002Fgithub.com\u002Fwuhuikai\u002FFastFCN)[ax1903][[project]](http:\u002F\u002Fwuhuikai.me\u002FFastFCNProject\u002F) [pytorch]\n\n\u003Ca id=\"framework_s__8\">\u003C\u002Fa>\n### Frameworks\n+ [ OpenMMLab Semantic Segmentation Toolbox and Benchmark](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmsegmentation) [pytorch]\n\n\u003Ca id=\"polyp_\">\u003C\u002Fa>\n### Polyp\n\n* [PraNet: Parallel Reverse Attention Network for Polyp Segmentation](https:\u002F\u002Fgithub.com\u002FDengPingFan\u002FPraNet)[miccai20]\n* [PHarDNet-MSEG: A Simple Encoder-Decoder Polyp Segmentation Neural Network that Achieves over 0.9 Mean Dice and 86 FPS](https:\u002F\u002Fgithub.com\u002Fjames128333\u002FHarDNet-MSEG)[ax2101]\n\n\u003Ca id=\"panoptic_segmentation_\">\u003C\u002Fa>\n## Panoptic Segmentation\n* [Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetectron2\u002Ftree\u002Fmain\u002Fprojects\u002FPanoptic-DeepLab) [cvpr20] [pytorch]\n\n\u003Ca id=\"video_segmentation__1\">\u003C\u002Fa>\n## Video Segmentation\n* [Improving Semantic Segmentation via Video Prediction and Label Relaxation](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fsemantic-segmentation) [cvpr19] [pytorch] [nvidia]\n* [PReMVOS: Proposal-generation, Refinement and Merging for Video Object Segmentation](https:\u002F\u002Fgithub.com\u002FJonathonLuiten\u002FPReMVOS) [accv18\u002Fcvprw18\u002Feccvw18] [tensorflow]\n* [MaskTrackRCNN for video instance segmentation](https:\u002F\u002Fgithub.com\u002Fyoutubevos\u002FMaskTrackRCNN) [iccv19] [pytorch\u002Fdetectron]\n* [MaskTrackRCNN](https:\u002F\u002Fgithub.com\u002Fyoutubevos\u002FMaskTrackRCNN) [iccv19] [pytorch\u002Fdetectron]\n* [Video Instance Segmentation using Inter-Frame Communication Transformers](https:\u002F\u002Fgithub.com\u002Fsukjunhwang\u002FIFC) [nips21] [pytorch\u002Fdetectron]\n* [VNext: SeqFormer \u002F IDOL](https:\u002F\u002Fgithub.com\u002Fwjf5203\u002FVNext) [eccv22] [pytorch\u002Fdetectron2]\n* [SeqFormer: Sequential Transformer for Video Instance Segmentation](https:\u002F\u002Fgithub.com\u002Fwjf5203\u002FSeqFormer) [eccv22] [pytorch\u002Fdetectron2]\n* [VITA: Video Instance Segmentation via Object Token Association](https:\u002F\u002Fgithub.com\u002Fsukjunhwang\u002Fvita) [nips22] [pytorch\u002Fdetectron2]\n\n\u003Ca id=\"panoptic_video_segmentation_\">\u003C\u002Fa>\n### Panoptic Video Segmentation\n* [ViP-DeepLab](https:\u002F\u002Fgithub.com\u002Fjoe-siyuan-qiao\u002FViP-DeepLab) [cvpr21] \n\n\u003Ca id=\"motion_prediction__1\">\u003C\u002Fa>\n## Motion Prediction\n* [Self-Supervised Learning via Conditional Motion Propagation](https:\u002F\u002Fgithub.com\u002FXiaohangZhan\u002Fconditional-motion-propagation) [cvpr19] [pytorch]\n* [A Neural Temporal Model for Human Motion Prediction](https:\u002F\u002Fgithub.com\u002Fcr7anand\u002Fneural_temporal_models) [cvpr19] [tensorflow]   \n* [Learning Trajectory Dependencies for Human Motion Prediction](https:\u002F\u002Fgithub.com\u002Fwei-mao-2019\u002FLearnTrajDep) [iccv19] [pytorch]   \n* [Structural-RNN: Deep Learning on Spatio-Temporal Graphs](https:\u002F\u002Fgithub.com\u002Fzhaolongkzz\u002Fhuman_motion) [cvpr15] [tensorflow]   \n* [A Keras multi-input multi-output LSTM-based RNN for object trajectory forecasting](https:\u002F\u002Fgithub.com\u002FMarlonCajamarca\u002FKeras-LSTM-Trajectory-Prediction) [keras]   \n* [Transformer Networks for Trajectory Forecasting](https:\u002F\u002Fgithub.com\u002FFGiuliari\u002FTrajectory-Transformer) [ax2003] [pytorch]  \n* [Regularizing neural networks for future trajectory prediction via IRL framework](https:\u002F\u002Fgithub.com\u002Fd1024choi\u002Ftraj-pred-irl) [ietcv1907] [tensorflow]  \n* [Peeking into the Future: Predicting Future Person Activities and Locations in Videos](https:\u002F\u002Fgithub.com\u002FJunweiLiang\u002Fnext-prediction) [cvpr19] [tensorflow]  \n* [DAG-Net: Double Attentive Graph Neural Network for Trajectory Forecasting](https:\u002F\u002Fgithub.com\u002Falexmonti19\u002Fdagnet) [ax200526] [pytorch]  \n* [MCENET: Multi-Context Encoder Network for Homogeneous Agent Trajectory Prediction in Mixed Traffic](https:\u002F\u002Fgithub.com\u002Fsugmichaelyang\u002FMCENET) [ax200405] [tensorflow]  \n* [Human Trajectory Prediction in Socially Interacting Crowds Using a CNN-based Architecture](https:\u002F\u002Fgithub.com\u002Fbiy001\u002Fsocial-cnn-pytorch) [pytorch]  \n* [A tool set for trajectory prediction, ready for pip install](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FTrajectory-Prediction-Tools) [icai19\u002Fwacv19]  [pytorch]  \n* [RobustTP: End-to-End Trajectory Prediction for Heterogeneous Road-Agents in Dense Traffic with Noisy Sensor Inputs](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FTrajectory-Prediction-Tools) [acmcscs19]  [pytorch\u002Ftensorflow]  \n* [The Garden of Forking Paths: Towards Multi-Future Trajectory Prediction](https:\u002F\u002Fgithub.com\u002FJunweiLiang\u002FMultiverse) [cvpr20] [dummy] \n* [Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Prediction](https:\u002F\u002Fgithub.com\u002Flmb-freiburg\u002FMultimodal-Future-Prediction) [cvpr19] [tensorflow] \n* [Adversarial Loss for Human Trajectory Prediction](https:\u002F\u002Fgithub.com\u002Fvita-epfl\u002FAdversarialLoss-SGAN) [hEART19] [pytorch] \n* [Social GAN: SSocially Acceptable Trajectories with Generative Adversarial Networks](https:\u002F\u002Fgithub.com\u002Fagrimgupta92\u002Fsgan) [cvpr18] [pytorch] \n* [Forecasting Trajectory and Behavior of Road-Agents Using Spectral Clustering in Graph-LSTMs](https:\u002F\u002Fgithub.com\u002Frohanchandra30\u002FSpectral-Trajectory-and-Behavior-Prediction) [ax1912] [pytorch] \n* [Study of attention mechanisms for trajectory prediction in Deep Learning](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FTrajectory-Prediction-Tools) [msc thesis]  [python]  \n* [A python implementation of multi-model estimation algorithm for trajectory tracking and prediction, research project from BMW ABSOLUT self-driving bus project.](https:\u002F\u002Fgithub.com\u002FchrisHuxi\u002FTrajectory_Predictor) [python]  \n* [Prediciting Human Trajectories](https:\u002F\u002Fgithub.com\u002Fkarthik4444\u002Fnn-trajectory-prediction) [theano]  \n* [Implementation of Recurrent Neural Networks for future trajectory prediction of pedestrians](https:\u002F\u002Fgithub.com\u002Faroongta\u002FPedestrian_Trajectory_Prediction) [pytorch]  \n\n\u003Ca id=\"pose_estimation_\">\u003C\u002Fa>\n## Pose Estimation\n\u003Ca id=\"framework_s__9\">\u003C\u002Fa>\n### Frameworks\n+ [OpenMMLab Pose Estimation Toolbox and Benchmark. ](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmpose) [pytorch]\n\n\u003Ca id=\"autoencoder_s_\">\u003C\u002Fa>\n## Autoencoders\n* [β-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework](https:\u002F\u002Fopenreview.net\u002Fforum?id=Sy2fzU9gl) [iclr17] [deepmind] [[tensorflow]](https:\u002F\u002Fgithub.com\u002Fmiyosuda\u002Fdisentangled_vae) [[tensorflow]](https:\u002F\u002Fgithub.com\u002FLynnHo\u002FVAE-Tensorflow) [[pytorch]](https:\u002F\u002Fgithub.com\u002F1Konny\u002FBeta-VAE)\n* [Disentangling by Factorising](https:\u002F\u002Fgithub.com\u002F1Konny\u002FFactorVAE) [ax1806] [pytorch]   \n\n\u003Ca id=\"classificatio_n__1\">\u003C\u002Fa>\n## Classification\n* [Learning Efficient Convolutional Networks Through Network Slimming](https:\u002F\u002Fgithub.com\u002Fmiyosuda\u002Fasync_deep_reinforce) [iccv17] [pytorch]\n\n\u003Ca id=\"framework_s__10\">\u003C\u002Fa>\n### Frameworks\n+ [OpenMMLab Image Classification Toolbox and Benchmark](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmclassification) [pytorch]\n\n\u003Ca id=\"deep_rl_\">\u003C\u002Fa>\n## Deep RL\n* [Asynchronous Methods for Deep Reinforcement Learning ](https:\u002F\u002Fgithub.com\u002Fmiyosuda\u002Fasync_deep_reinforce)\n\n\u003Ca id=\"annotatio_n_\">\u003C\u002Fa>\n## Annotation\n- [LabelImg](https:\u002F\u002Fgithub.com\u002Ftzutalin\u002FlabelImg)\n- [ByLabel: A Boundary Based Semi-Automatic Image Annotation Tool](https:\u002F\u002Fgithub.com\u002FNathanUA\u002FByLabel)\n- [Bounding Box Editor and Exporter](https:\u002F\u002Fgithub.com\u002Fpersts\u002FBBoxEE)\n- [VGG Image Annotator](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~vgg\u002Fsoftware\u002Fvia\u002F)\n- [Visual Object Tagging Tool: An electron app for building end to end Object Detection Models from Images and Videos](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002FVoTT)\n- [PixelAnnotationTool](https:\u002F\u002Fgithub.com\u002Fabreheret\u002FPixelAnnotationTool)\n- [labelme : Image Polygonal Annotation with Python (polygon, rectangle, circle, line, point and image-level flag annotation)](https:\u002F\u002Fgithub.com\u002Fwkentaro\u002Flabelme)\n- [VATIC - Video Annotation Tool from Irvine, California)](https:\u002F\u002Fgithub.com\u002Fcvondrick\u002Fvatic) [ijcv12] [[project]](http:\u002F\u002Fwww.cs.columbia.edu\u002F~vondrick\u002Fvatic\u002F)\n- [Computer Vision Annotation Tool (CVAT)](https:\u002F\u002Fgithub.com\u002Fopencv\u002Fcvat)\n- [Image labelling tool](https:\u002F\u002Fbitbucket.org\u002Fueacomputervision\u002Fimage-labelling-tool\u002F)\n- [Labelbox](https:\u002F\u002Fgithub.com\u002FLabelbox\u002FLabelbox) [paid]\n- [RectLabel An image annotation tool to label images for bounding box object detection and segmentation.](https:\u002F\u002Frectlabel.com\u002F) [paid]\n- [Onepanel: Production scale vision AI platform with fully integrated components for model building, automated labeling, data processing and model training pipelines.](https:\u002F\u002Fgithub.com\u002Fonepanelio\u002Fcore) [[docs]](https:\u002F\u002Fdocs.onepanel.ai\u002Fdocs\u002Fgetting-started\u002Fquickstart\u002F)\n\n\u003Ca id=\"editing_\">\u003C\u002Fa>\n### Editing\n- [OpenMMLab Image and Video Editing Toolbox](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmediting)\n\n\u003Ca id=\"augmentatio_n_\">\u003C\u002Fa>\n### Augmentation\n- [Augmentor: Image augmentation library in Python for machine learning](https:\u002F\u002Fgithub.com\u002Fmdbloice\u002FAugmentor)\n- [Albumentations: Fast image augmentation library and easy to use wrapper around other libraries](https:\u002F\u002Fgithub.com\u002Falbumentations-team\u002Falbumentations)\n- [imgaug: Image augmentation for machine learning experiments](https:\u002F\u002Fgithub.com\u002Faleju\u002Fimgaug)\n- [solt: Image Streaming over lightweight data transformations](https:\u002F\u002Fgithub.com\u002FMIPT-Oulu\u002Fsolt)\n\n\u003Ca id=\"deep_learning__2\">\u003C\u002Fa>\n## Deep Learning\n* [Deformable Convolutional Networks](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FDeformable-ConvNets)\n* [RNNexp](https:\u002F\u002Fgithub.com\u002Fasheshjain399\u002FRNNexp)\n* [Grad-CAM: Gradient-weighted Class Activation Mapping](https:\u002F\u002Fgithub.com\u002Framprs\u002Fgrad-cam\u002F)\n\n\u003Ca id=\"class_imbalanc_e_\">\u003C\u002Fa>\n### Class Imbalance\n* [Imbalanced Dataset Sampler](https:\u002F\u002Fgithub.com\u002Fufoym\u002Fimbalanced-dataset-sampler) [pytorch]\n* [Iterable dataset resampling in PyTorch](https:\u002F\u002Fgithub.com\u002FMaxHalford\u002Fpytorch-resample) [pytorch]\n\n\u003Ca id=\"few_shot_learning_\">\u003C\u002Fa>\n### Few shot learning\n* [OpenMMLab FewShot Learning Toolbox and Benchmark](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmfewshot) [pytorch]\n\n\u003Ca id=\"unsupervised_learning__2\">\u003C\u002Fa>\n### Unsupervised learning\n* [Self-Supervised Learning Toolbox and Benchmark](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002FOpenSelfSup) [pytorch]\n\n\u003Ca id=\"collections_\">\u003C\u002Fa>\n# Collections\n\n\u003Ca id=\"dataset_s__1\">\u003C\u002Fa>\n## Datasets\n* [Awesome Public Datasets](https:\u002F\u002Fgithub.com\u002Fawesomedata\u002Fawesome-public-datasets) \n* [List of traffic surveillance datasets](https:\u002F\u002Fgithub.com\u002Fgustavovelascoh\u002Ftraffic-surveillance-dataset) \n* [Machine learning datasets: A list of the biggest machine learning datasets from across the web](https:\u002F\u002Fwww.datasetlist.com\u002F) \n* [Labeled Information Library of Alexandria: Biology and Conservation](http:\u002F\u002Flila.science\u002Fdatasets) [[other conservation data sets]](http:\u002F\u002Flila.science\u002Fotherdatasets) \n* [THOTH: Data Sets & Images](https:\u002F\u002Fthoth.inrialpes.fr\u002Fdata) \n* [Google AI Datasets](https:\u002F\u002Fai.google\u002Ftools\u002Fdatasets\u002F) \n* [Google Cloud Storage public datasets](https:\u002F\u002Fcloud.google.com\u002Fstorage\u002Fdocs\u002Fpublic-datasets\u002F) \n* [Microsoft Research Open Data](https:\u002F\u002Fmsropendata.com\u002F) \n* [Earth Engine Data Catalog](https:\u002F\u002Fdevelopers.google.com\u002Fearth-engine\u002Fdatasets\u002Fcatalog\u002F) \n* [Registry of Open Data on AWS](https:\u002F\u002Fregistry.opendata.aws\u002F) \n* [Kaggle Datasets](https:\u002F\u002Fwww.kaggle.com\u002Fdatasets) \n* [CVonline: Image Databases](http:\u002F\u002Fhomepages.inf.ed.ac.uk\u002Frbf\u002FCVonline\u002FImagedbase.htm) \n* [Synthetic for Computer Vision: A list of synthetic dataset and tools for computer vision](https:\u002F\u002Fgithub.com\u002Funrealcv\u002Fsynthetic-computer-vision) \n* [pgram machine learning datasets](https:\u002F\u002Fpgram.com\u002Fcategory\u002Fvision\u002F) \n* [pgram vision datasets](https:\u002F\u002Fpgram.com\u002F) \n\n\u003Ca id=\"deep_learning__3\">\u003C\u002Fa>\n## Deep Learning\n- [Model Zoo : Discover open source deep learning code and pretrained models](https:\u002F\u002Fmodelzoo.co\u002F)\n\n\u003Ca id=\"static_detectio_n__2\">\u003C\u002Fa>\n## Static Detection\n- [Object Detection with Deep Learning](https:\u002F\u002Fhandong1587.github.io\u002Fdeep_learning\u002F2015\u002F10\u002F09\u002Fobject-detection.html)\n\n\u003Ca id=\"video_detectio_n__3\">\u003C\u002Fa>\n## Video Detection\n- [Video Object Detection with Deep Learning](https:\u002F\u002Fhandong1587.github.io\u002Fdeep_learning\u002F2015\u002F10\u002F09\u002Fobject-detection.html#video-object-detection)\n\n\u003Ca id=\"single_object_tracking__3\">\u003C\u002Fa>\n## Single Object Tracking\n- [Visual Tracking Paper List](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002Fbenchmark_results)\n- [List of deep learning based tracking papers](https:\u002F\u002Fgithub.com\u002Fhandong1587\u002Fhandong1587.github.io\u002Fblob\u002Fmaster\u002F_posts\u002Fdeep_learning\u002F2015-10-09-tracking.md)\n- [List of single object trackers with results on OTB](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002Fbenchmark_results)\n- [Collection of Correlation Filter based trackers with links to papers, codes, etc](https:\u002F\u002Fgithub.com\u002Flukaswals\u002Fcf-trackers)\n- [VOT2018 Trackers repository](http:\u002F\u002Fwww.votchallenge.net\u002Fvot2018\u002Ftrackers.html)\n- [CUHK Datasets](http:\u002F\u002Fmmlab.ie.cuhk.edu.hk.login.ezproxy.library.ualberta.ca\u002Fdatasets.html)\n- [A Summary of CVPR19 Visual Tracking Papers](https:\u002F\u002Flinkinpark213.com\u002F2019\u002F06\u002F11\u002Fcvpr19-track\u002F )\n- [Visual Trackers for Single Object](https:\u002F\u002Fgithub.com\u002Fczla\u002Fdaily-paper-visual-tracking)\n    \n\u003Ca id=\"multi_object_tracking__3\">\u003C\u002Fa>\n## Multi Object Tracking\n* [List of multi object tracking papers](http:\u002F\u002Fperception.yale.edu\u002FBrian\u002FrefGuides\u002FMOT.html)   \n* [A collection of Multiple Object Tracking (MOT) papers in recent years, with notes](https:\u002F\u002Fgithub.com\u002Fhuanglianghua\u002Fmot-papers)  \n* [Papers with Code : Multiple Object Tracking](https:\u002F\u002Fpaperswithcode.com\u002Ftask\u002Fmultiple-object-tracking\u002Fcodeless)  \n* [Paper list and source code for multi-object-tracking](https:\u002F\u002Fgithub.com\u002FSpyderXu\u002Fmulti-object-tracking-paper-list)  \n\n\u003Ca id=\"static_segmentation__1\">\u003C\u002Fa>\n## Static Segmentation\n* [Segmentation Papers and Code](https:\u002F\u002Fhandong1587.github.io\u002Fdeep_learning\u002F2015\u002F10\u002F09\u002Fsegmentation.html)  \n* [Segmentation.X : Papers and Benchmarks about semantic segmentation, instance segmentation, panoptic segmentation and video segmentation](https:\u002F\u002Fgithub.com\u002FwutianyiRosun\u002FSegmentation.X) \n* [Instance Segmentation Papers with Code](https:\u002F\u002Fpaperswithcode.com\u002Ftask\u002Finstance-segmentation) \n\n\u003Ca id=\"video_segmentation__2\">\u003C\u002Fa>\n## Video Segmentation\n* [Video Instance Segmentation on YouTube-VIS validation](https:\u002F\u002Fpaperswithcode.com\u002Fsota\u002Fvideo-instance-segmentation-on-youtube-vis-1?p=seqformer-a-frustratingly-simple-model-for)\n\n\n\u003Ca id=\"motion_prediction__2\">\u003C\u002Fa>\n## Motion Prediction\n* [Awesome-Trajectory-Prediction](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FAwesome-Trajectory-Prediction\u002Fblob\u002Fmaster\u002FREADME.md)  \n* [Awesome Interaction-aware Behavior and Trajectory Prediction](https:\u002F\u002Fgithub.com\u002Fjiachenli94\u002FAwesome-Interaction-aware-Trajectory-Prediction)  \n* [Human Trajectory Prediction Datasets](https:\u002F\u002Fgithub.com\u002Famiryanj\u002FOpenTraj)  \n\n\u003Ca id=\"deep_compressed_sensin_g_\">\u003C\u002Fa>\n## Deep Compressed Sensing\n* [Reproducible Deep Compressive Sensing](https:\u002F\u002Fgithub.com\u002FAtenaKid\u002FReproducible-Deep-Compressive-Sensing)  \n\n\u003Ca id=\"mis_c__3\">\u003C\u002Fa>\n## Misc \n* [Papers With Code : the latest in machine learning](https:\u002F\u002Fpaperswithcode.com\u002F)\n* [Awesome Deep Ecology](https:\u002F\u002Fgithub.com\u002Fpatrickcgray\u002Fawesome-deep-ecology)\n* [List of Matlab frameworks, libraries and software](https:\u002F\u002Fgithub.com\u002Fuhub\u002Fawesome-matlab)\n* [Face Recognition](https:\u002F\u002Fgithub.com\u002FChanChiChoi\u002Fawesome-Face_Recognition)\n* [A Month of Machine Learning Paper Summaries](https:\u002F\u002Fmedium.com\u002F@hyponymous\u002Fa-month-of-machine-learning-paper-summaries-ddd4dcf6cfa5)\n* [Awesome-model-compression-and-acceleration](https:\u002F\u002Fgithub.com\u002Fmemoiry\u002FAwesome-model-compression-and-acceleration\u002Fblob\u002Fmaster\u002FREADME.md)\n* [Model-Compression-Papers](https:\u002F\u002Fgithub.com\u002Fchester256\u002FModel-Compression-Papers)\n\n\u003Ca id=\"tutorials_\">\u003C\u002Fa>\n# Tutorials\n\n\u003Ca id=\"collections__1\">\u003C\u002Fa>\n## Collections\n* [Deep Tutorials for PyTorch](https:\u002F\u002Fgithub.com\u002Fsgrvinod\u002FDeep-Tutorials-for-PyTorch)\n\n\u003Ca id=\"multi_object_tracking__4\">\u003C\u002Fa>\n## Multi Object Tracking\n* [What is the Multi-Object Tracking (MOT) system?](https:\u002F\u002Fdeepomatic.com\u002Fen\u002Fmoving-beyond-deepomatic-learns-how-to-track-multiple-objects\u002F)\n\n\u003Ca id=\"static_detectio_n__3\">\u003C\u002Fa>\n## Static Detection\n* [End-to-end object detection with Transformers](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fend-to-end-object-detection-with-transformers)\n* [Deep Learning for Object Detection: A Comprehensive Review](https:\u002F\u002Ftowardsdatascience.com\u002Fdeep-learning-for-object-detection-a-comprehensive-review-73930816d8d9)\n* [Review of Deep Learning Algorithms for Object Detection](https:\u002F\u002Fmedium.com\u002Fcomet-app\u002Freview-of-deep-learning-algorithms-for-object-detection-c1f3d437b852)  \n* [A Simple Guide to the Versions of the Inception Network](https:\u002F\u002Ftowardsdatascience.com\u002Fa-simple-guide-to-the-versions-of-the-inception-network-7fc52b863202)    \n* [R-CNN, Fast R-CNN, Faster R-CNN, YOLO - Object Detection Algorithms](https:\u002F\u002Ftowardsdatascience.com\u002Fr-cnn-fast-r-cnn-faster-r-cnn-yolo-object-detection-algorithms-36d53571365e)\n* [A gentle guide to deep learning object detection](https:\u002F\u002Fwww.pyimagesearch.com\u002F2018\u002F05\u002F14\u002Fa-gentle-guide-to-deep-learning-object-detection\u002F)\n* [The intuition behind RetinaNet](https:\u002F\u002Fmedium.com\u002F@14prakash\u002Fthe-intuition-behind-retinanet-eb636755607d)\n* [YOLO—You only look once, real time object detection explained](https:\u002F\u002Ftowardsdatascience.com\u002Fyolo-you-only-look-once-real-time-object-detection-explained-492dc9230006)\n* [Understanding Feature Pyramid Networks for object detection (FPN)](https:\u002F\u002Fmedium.com\u002F@jonathan_hui\u002Funderstanding-feature-pyramid-networks-for-object-detection-fpn-45b227b9106c)\n* [Fast object detection with SqueezeDet on Keras](https:\u002F\u002Fmedium.com\u002Fomnius\u002Ffast-object-detection-with-squeezedet-on-keras-5cdd124b46ce)\n* [Region of interest pooling explained](https:\u002F\u002Fdeepsense.ai\u002Fregion-of-interest-pooling-explained\u002F)\n\n\u003Ca id=\"video_detectio_n__4\">\u003C\u002Fa>\n## Video Detection\n* [How Microsoft Does Video Object Detection - Unifying the Best Techniques in Video Object Detection Architectures in a Single Model](https:\u002F\u002Fmedium.com\u002Fnurture-ai\u002Fhow-microsoft-does-video-object-detection-unifying-the-best-techniques-in-video-object-detection-b78b63e3f1d8)\n\n\u003Ca id=\"instance_segmentation__1\">\u003C\u002Fa>\n## Instance Segmentation\n* [Splash of Color: Instance Segmentation with Mask R-CNN and TensorFlow](https:\u002F\u002Fengineering.matterport.com\u002Fsplash-of-color-instance-segmentation-with-mask-r-cnn-and-tensorflow-7c761e238b46)\n* [Simple Understanding of Mask RCNN](https:\u002F\u002Fmedium.com\u002F@alittlepain833\u002Fsimple-understanding-of-mask-rcnn-134b5b330e95)\n* [Learning to Segment](https:\u002F\u002Fresearch.fb.com\u002Fblog\u002F2016\u002F08\u002Flearning-to-segment\u002F)\n* [Analyzing The Papers Behind Facebook's Computer Vision Approach](https:\u002F\u002Fadeshpande3.github.io\u002FAnalyzing-the-Papers-Behind-Facebook's-Computer-Vision-Approach\u002F)\n* [Review: MNC — Multi-task Network Cascade, Winner in 2015 COCO Segmentation](https:\u002F\u002Ftowardsdatascience.com\u002Freview-mnc-multi-task-network-cascade-winner-in-2015-coco-segmentation-instance-segmentation-42a9334e6a34)\n* [Review: FCIS — Winner in 2016 COCO Segmentation](https:\u002F\u002Ftowardsdatascience.com\u002Freview-fcis-winner-in-2016-coco-segmentation-instance-segmentation-ee2d61f465e2)\n* [Review: InstanceFCN — Instance-Sensitive Score Maps](https:\u002F\u002Ftowardsdatascience.com\u002Freview-instancefcn-instance-sensitive-score-maps-instance-segmentation-dbfe67d4ee92)\n\n\u003Ca id=\"deep_learning__4\">\u003C\u002Fa>\n## Deep Learning\n\n\u003Ca id=\"optimizatio_n_\">\u003C\u002Fa>\n### Optimization\n* [Learning Rate Scheduling](https:\u002F\u002Fwww.deeplearningwizard.com\u002Fdeep_learning\u002Fboosting_models_pytorch\u002Flr_scheduling\u002F)\n\n\u003Ca id=\"class_imbalanc_e__1\">\u003C\u002Fa>\n### Class Imbalance\n* [Learning from imbalanced data](https:\u002F\u002Fwww.jeremyjordan.me\u002Fimbalanced-data\u002F)\n* [Learning from Imbalanced Classes](https:\u002F\u002Fwww.svds.com\u002Flearning-imbalanced-classes\u002F)\n* [Handling imbalanced datasets in machine learning](https:\u002F\u002Ftowardsdatascience.com\u002Fhandling-imbalanced-datasets-in-machine-learning-7a0e84220f28) [medium]\n* [How to handle Class Imbalance Problem](https:\u002F\u002Fmedium.com\u002Fquantyca\u002Fhow-to-handle-class-imbalance-problem-9ee3062f2499) [medium]\n* [Dealing with Imbalanced Data](https:\u002F\u002Ftowardsdatascience.com\u002Fmethods-for-dealing-with-imbalanced-data-5b761be45a18) [towardsdatascience]\n* [How to Handle Imbalanced Classes in Machine Learning](https:\u002F\u002Felitedatascience.com\u002Fimbalanced-classes) [elitedatascience]\n* [7 Techniques to Handle Imbalanced Data](https:\u002F\u002Fwww.kdnuggets.com\u002F2017\u002F06\u002F7-techniques-handle-imbalanced-data.html) [kdnuggets]\n* [10 Techniques to deal with Imbalanced Classes in Machine Learning](https:\u002F\u002Fwww.analyticsvidhya.com\u002Fblog\u002F2020\u002F07\u002F10-techniques-to-deal-with-class-imbalance-in-machine-learning\u002F) [analyticsvidhya]\n\n\u003Ca id=\"rnn__2\">\u003C\u002Fa>\n## RNN\n* [The Unreasonable Effectiveness of Recurrent Neural Networks](http:\u002F\u002Fkarpathy.github.io\u002F2015\u002F05\u002F21\u002Frnn-effectiveness\u002F)\n* [Understanding LSTM Networks](http:\u002F\u002Fcolah.github.io\u002Fposts\u002F2015-08-Understanding-LSTMs\u002F)\n\n\u003Ca id=\"deep_rl__1\">\u003C\u002Fa>\n## Deep RL\n* [Deep Reinforcement Learning: Pong from Pixels](http:\u002F\u002Fkarpathy.github.io\u002F2016\u002F05\u002F31\u002Frl\u002F)\n* [Demystifying Deep Reinforcement Learning](https:\u002F\u002Fwww.intelnervana.com\u002Fdemystifying-deep-reinforcement-learning\u002F)\n\n\u003Ca id=\"autoencoder_s__1\">\u003C\u002Fa>\n## Autoencoders\n* [Guide to Autoencoders](https:\u002F\u002Fyaledatascience.github.io\u002F2016\u002F10\u002F29\u002Fautoencoders.html)\n* [Applied Deep Learning - Part 3: Autoencoders](https:\u002F\u002Ftowardsdatascience.com\u002Fapplied-deep-learning-part-3-autoencoders-1c083af4d798)\n* [Denoising Autoencoders](http:\u002F\u002Fdeeplearning.net\u002Ftutorial\u002FdA.html)\n* [Stacked Denoising Autoencoders](https:\u002F\u002Fskymind.ai\u002Fwiki\u002Fstacked-denoising-autoencoder)\n* [A Gentle Introduction to LSTM Autoencoders](https:\u002F\u002Fmachinelearningmastery.com\u002Flstm-autoencoders\u002F)\n* [Variational Autoencoder in TensorFlow](https:\u002F\u002Fjmetzen.github.io\u002F2015-11-27\u002Fvae.html)\n* [Variational Autoencoders with Tensorflow Probability Layers](https:\u002F\u002Fmedium.com\u002Ftensorflow\u002Fvariational-autoencoders-with-tensorflow-probability-layers-d06c658931b7)\n\n\n\u003Ca id=\"blogs_\">\u003C\u002Fa>\n# Blogs\n\n* [Facebook AI](https:\u002F\u002Fai.facebook.com\u002Fblog\u002F)\n* [Google AI](https:\u002F\u002Fai.googleblog.com\u002F)\n* [Google DeepMind](https:\u002F\u002Fdeepmind.com\u002Fblog)\n* [Deep Learning Wizard](https:\u002F\u002Fwww.deeplearningwizard.com\u002F)\n* [Towards Data Science](https:\u002F\u002Ftowardsdatascience.com\u002F)\n* [Jay Alammar : Visualizing machine learning one concept at a time](https:\u002F\u002Fjalammar.github.io\u002F)\n* [Inside Machine Learning: Deep-dive articles about machine learning, cloud, and data. Curated by IBM](https:\u002F\u002Fmedium.com\u002Finside-machine-learning)\n* [colah's blog](http:\u002F\u002Fcolah.github.io\u002F)\n* [Jeremy Jordan](https:\u002F\u002Fwww.jeremyjordan.me\u002F)\n* [Silicon Valley Data Science](https:\u002F\u002Fwww.svds.com\u002Ftag\u002Fdata-science\u002F)\n* [Illarion’s Notes](https:\u002F\u002Fikhlestov.github.io\u002Fpages\u002F)\n\n","\u003C!-- No Heading Fix -->\n使用深度学习进行目标检测与跟踪的论文、数据集、代码及其他资源合集\n\u003C!-- MarkdownTOC -->\n\n- [研究数据](#research_data_)\n- [论文](#paper_s_)\n    - [静态检测](#static_detectio_n_)\n        - [区域提议（Region Proposal）](#region_proposal_)\n        - [RCNN](#rcn_n_)\n        - [YOLO](#yol_o_)\n        - [SSD](#ssd_)\n        - [RetinaNet](#retinanet_)\n        - [无锚框（Anchor Free）](#anchor_free_)\n        - [其他](#mis_c_)\n    - [视频检测](#video_detectio_n_)\n        - [Tubelet](#tubelet_)\n        - [FGFA](#fgf_a_)\n        - [RNN](#rnn_)\n    - [多目标跟踪（Multi Object Tracking）](#multi_object_tracking_)\n        - [联合检测（Joint-Detection）](#joint_detection_)\n            - [身份嵌入（Identity Embedding）](#identity_embeddin_g_)\n        - [关联（Association）](#association_)\n        - [深度学习](#deep_learning_)\n        - [RNN](#rnn__1)\n        - [无监督学习（Unsupervised Learning）](#unsupervised_learning_)\n        - [强化学习（Reinforcement Learning）](#reinforcement_learning_)\n        - [网络流（Network Flow）](#network_flow_)\n        - [图优化（Graph Optimization）](#graph_optimization_)\n        - [基线方法（Baseline）](#baselin_e_)\n        - [评估指标（Metrics）](#metrics_)\n    - [单目标跟踪（Single Object Tracking）](#single_object_tracking_)\n        - [强化学习](#reinforcement_learning__1)\n        - [孪生网络（Siamese）](#siamese_)\n        - [相关滤波（Correlation）](#correlation_)\n        - [其他](#mis_c__1)\n    - [深度学习](#deep_learning__1)\n        - [合成梯度（Synthetic Gradients）](#synthetic_gradient_s_)\n        - [高效模型（Efficient）](#efficient_)\n    - [无监督学习](#unsupervised_learning__1)\n    - [插值（Interpolation）](#interpolation_)\n    - [自编码器（Autoencoder）](#autoencoder_)\n        - [变分自编码器（Variational）](#variational_)\n- [数据集](#dataset_s_)\n    - [多目标跟踪](#multi_object_tracking__1)\n        - [无人机（UAV）](#uav_)\n        - [合成数据（Synthetic）](#synthetic_)\n        - [显微镜 \u002F 细胞跟踪（Microscopy \u002F Cell Tracking）](#microscopy___cell_tracking_)\n    - [单目标跟踪](#single_object_tracking__1)\n    - [视频检测](#video_detectio_n__1)\n        - [视频理解 \u002F 动作识别（Video Understanding \u002F Activity Recognition）](#video_understanding___activity_recognitio_n_)\n    - [静态检测](#static_detectio_n__1)\n        - [动物（Animals）](#animals_)\n    - [边界检测（Boundary Detection）](#boundary_detectio_n_)\n    - [静态分割（Static Segmentation）](#static_segmentation_)\n    - [视频分割（Video Segmentation）](#video_segmentation_)\n    - [分类（Classification）](#classificatio_n_)\n    - [光流（Optical Flow）](#optical_flow_)\n    - [运动预测（Motion Prediction）](#motion_prediction_)\n- [代码](#cod_e_)\n    - [通用视觉（General Vision）](#general_vision_)\n    - [多目标跟踪](#multi_object_tracking__2)\n        - [框架（Frameworks）](#framework_s_)\n        - [通用工具](#general_)\n        - [基线方法](#baselin_e__1)\n        - [孪生网络](#siamese__1)\n        - [无监督方法](#unsupervise_d_)\n        - [行人重识别（Re-ID）](#re_id_)\n            - [框架](#framework_s__1)\n        - [图神经网络（Graph NN）](#graph_nn_)\n        - [显微镜 \u002F 细胞跟踪](#microscopy___cell_tracking__1)\n        - [3D](#3_d_)\n        - [评估指标](#metrics__1)\n    - [单目标跟踪](#single_object_tracking__2)\n        - [GUI 应用 \u002F 大规模跟踪 \u002F 动物](#gui_application___large_scale_tracking___animal_s_)\n    - [视频检测](#video_detectio_n__2)\n        - [动作检测（Action Detection）](#action_detectio_n_)\n            - [框架](#framework_s__2)\n    - [静态检测与匹配（Static Detection and Matching）](#static_detection_and_matching_)\n        - [框架](#framework_s__3)\n        - [区域提议](#region_proposal__1)\n        - [特征金字塔网络（FPN）](#fpn_)\n        - [RCNN](#rcn_n__1)\n        - [SSD](#ssd__1)\n        - [RetinaNet](#retinanet__1)\n        - [YOLO](#yol_o__1)\n        - [无锚框](#anchor_free__1)\n        - [其他](#mis_c__2)\n        - [匹配（Matching）](#matchin_g_)\n        - [边界检测](#boundary_detectio_n__1)\n        - [文本检测（Text Detection）](#text_detectio_n_)\n            - [框架](#framework_s__4)\n        - [3D 检测](#3d_detectio_n_)\n            - [框架](#framework_s__5)\n    - [光流](#optical_flow__1)\n        - [框架](#framework_s__6)\n    - [实例分割（Instance Segmentation）](#instance_segmentation_)\n        - [框架](#framework_s__7)\n    - [语义分割（Semantic Segmentation）](#semantic_segmentation_)\n        - [框架](#framework_s__8)\n        - [息肉（Polyp）](#polyp_)\n    - [全景分割（Panoptic Segmentation）](#panoptic_segmentation_)\n    - [视频分割](#video_segmentation__1)\n        - [全景视频分割（Panoptic Video Segmentation）](#panoptic_video_segmentation_)\n    - [运动预测](#motion_prediction__1)\n    - [姿态估计（Pose Estimation）](#pose_estimation_)\n        - [框架](#framework_s__9)\n    - [自编码器](#autoencoder_s_)\n    - [分类](#classificatio_n__1)\n        - [框架](#framework_s__10)\n    - [深度强化学习（Deep RL）](#deep_rl_)\n    - [标注（Annotation）](#annotatio_n_)\n        - [编辑（Editing）](#editing_)\n        - [数据增强（Augmentation）](#augmentatio_n_)\n    - [深度学习](#deep_learning__2)\n        - [类别不平衡（Class Imbalance）](#class_imbalanc_e_)\n        - [小样本学习（Few shot learning）](#few_shot_learning_)\n        - [无监督学习](#unsupervised_learning__2)\n- [资源合集](#collections_)\n    - [数据集](#dataset_s__1)\n    - [深度学习](#deep_learning__3)\n    - [静态检测](#static_detectio_n__2)\n    - [视频检测](#video_detectio_n__3)\n    - [单目标跟踪](#single_object_tracking__3)\n    - [多目标跟踪](#multi_object_tracking__3)\n    - [静态分割](#static_segmentation__1)\n    - [视频分割](#video_segmentation__2)\n    - [运动预测](#motion_prediction__2)\n    - [深度压缩感知（Deep Compressed Sensing）](#deep_compressed_sensin_g_)\n    - [其他](#mis_c__3)\n- [教程](#tutorials_)\n    - [资源合集](#collections__1)\n    - [多目标跟踪](#multi_object_tracking__4)\n    - [静态检测](#static_detectio_n__3)\n    - [视频检测](#video_detectio_n__4)\n    - [实例分割](#instance_segmentation__1)\n    - [深度学习](#deep_learning__4)\n        - [优化（Optimization）](#optimizatio_n_)\n        - [类别不平衡](#class_imbalanc_e__1)\n    - [RNN](#rnn__2)\n    - [深度强化学习](#deep_rl__1)\n    - [自编码器](#autoencoder_s__1)\n- [博客](#blogs_)\n\n\u003C!-- \u002FMarkdownTOC -->\n\n\u003Ca id=\"research_data_\">\u003C\u002Fa>\n\n# 研究数据\n\n我使用 [DavidRM Journal](http:\u002F\u002Fwww.davidrm.com\u002F) 来管理我的研究数据，因为它具备出色的层次化组织、交叉链接和标签（tagging）功能。\n\n我提供了一个 Journal 条目导出文件，其中包含我对过去几年收集的关于计算机视觉（computer vision）和深度学习（deep learning）的论文、文章、教程、代码和笔记进行分类和打标签后的整理成果。\n\n主题云（topic cloud）效果如下所示：  \n![Alt text](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fabhineet123_Deep-Learning-for-Tracking-and-Detection_readme_439592ee4cb1.jpg)\n\n该文件需要 Journal 8 版本，并可通过以下步骤导入：\n\n- 使用 **File** -> **Import** -> **Import User Preferences** 导入我的用户偏好设置\n- 使用 **File** -> **Import** -> **Sync from The Journal Export File** 导入研究数据\n\n请注意，必须先导入我的用户偏好设置，再导入研究数据，否则带标签的主题将无法正常工作。\n\n（可选）我还提供了全局选项（Global Options）文件，适用于喜欢深色主题（dark theme）的用户，可通过 **File** -> **Import** -> **Import Global Options** 导入。\n\n-  [用户偏好设置](research_data\u002Fuser_settings.juser)\n-  [条目导出文件](research_data\u002Fphd_literature_readings.tjexp)\n-  [全局选项](research_data\u002Fglobal_options.tjglobal)\n\n更新日期：2026-03-09\n\n\u003Ca id=\"paper_s_\">\u003C\u002Fa>\n# 论文\n\n\u003Ca id=\"static_detectio_n_\">\u003C\u002Fa>\n## 静态检测（Static Detection）\n\n\u003Ca id=\"region_proposal_\">\u003C\u002Fa>\n### 区域提议（Region Proposal）\n- **Scalable Object Detection Using Deep Neural Networks**  \n[cvpr14]  \n[[pdf]](static_detection\u002Fregion_proposal\u002FScalable%20Object%20Detection%20Using%20Deep%20Neural%20Networks%20cvpr14.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FScalable%20Object%20Detection%20Using%20Deep%20Neural%20Networks%20cvpr14.pdf)\n- **Selective Search for Object Recognition**  \n[ijcv2013]  \n[[pdf]](static_detection\u002Fregion_proposal\u002FSelective%20Search%20for%20Object%20Recognition%20ijcv2013.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FSelective%20Search%20for%20Object%20Recognition%20ijcv2013.pdf)\n\n\u003Ca id=\"rcn_n_\">\u003C\u002Fa>\n### RCNN\n- **Faster R-CNN Towards Real-Time Object Detection with Region Proposal Networks**  \n[tpami17]  \n[[pdf]](static_detection\u002FRCNN\u002FFaster%20R-CNN%20Towards%20Real-Time%20Object%20Detection%20with%20Region%20Proposal%20Networks%20tpami17%20ax16_1.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FFaster_R-CNN.pdf)\n- **RFCN - Object Detection via Region-based Fully Convolutional Networks**  \n[nips16]  \n[Microsoft Research]  \n[[pdf]](static_detection\u002FRCNN\u002FRFCN-Object%20Detection%20via%20Region-based%20Fully%20Convolutional%20Networks%20nips16.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FRFCN.pdf)  \n- **Mask R-CNN**  \n[iccv17]  \n[Facebook AI Research]  \n[[pdf]](static_detection\u002FRCNN\u002FMask%20R-CNN%20ax17_4%20iccv17.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FMask%20R-CNN%20ax17_4%20iccv17.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1703.06870)  \n[[code (keras)]](https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)  \n[[code (tensorflow)]](https:\u002F\u002Fgithub.com\u002FCharlesShang\u002FFastMaskRCNN)\n- **SNIPER Efficient Multi-Scale Training**  \n[ax1812\u002Fnips18]  \n[[pdf]](static_detection\u002FRCNN\u002FSNIPER%20Efficient%20Multi-Scale%20Training%20ax181213%20nips18.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FSNIPER%20Efficient%20Multi-Scale%20Training%20ax181213%20nips18.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fmahyarnajibi\u002FSNIPER)\n\n\n\u003Ca id=\"yol_o_\">\u003C\u002Fa>\n### YOLO\n- **You Only Look Once Unified, Real-Time Object Detection**  \n[ax1605]  \n[[pdf]](static_detection\u002Fyolo\u002FYou%20Only%20Look%20Once%20Unified,%20Real-Time%20Object%20Detection%20ax1605.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FYou%20Only%20Look%20Once%20Unified,%20Real-Time%20Object%20Detection%20ax1605.pdf)\n- **YOLO9000 Better, Faster, Stronger**  \n[ax1612]  \n[[pdf]](static_detection\u002Fyolo\u002FYOLO9000%20Better,%20Faster,%20Stronger%20ax16_12.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FYOLO9000%20Better,%20Faster,%20Stronger%20ax16_12.pdf)\n- **YOLOv3 An Incremental Improvement**  \n[ax1804]  \n[[pdf]](static_detection\u002Fyolo\u002FYOLOv3%20An%20Incremental%20Improvement%20ax180408.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FYOLOv3%20An%20Incremental%20Improvement%20ax180408.pdf)\n- **YOLOv4 Optimal Speed and Accuracy of Object Detection**  \n[ax2004]  \n[[pdf]](static_detection\u002Fyolo\u002FYOLOV4_Optimal%20Speed%20and%20Accuracy%20of%20Object%20Detection%20ax200423.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FYOLOV4_Optimal%20Speed%20and%20Accuracy%20of%20Object%20Detection%20ax200423.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002FAlexeyAB\u002Fdarknet)\n\n\u003Ca id=\"ssd_\">\u003C\u002Fa>\n### SSD\n- **SSD Single Shot MultiBox Detector**  \n[ax1612\u002Feccv16]  \n[[pdf]](static_detection\u002Fssd\u002FSSD%20Single%20Shot%20MultiBox%20Detector%20eccv16_ax16_12.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FSSD.pdf)\n- **DSSD  Deconvolutional Single Shot Detector**  \n[ax1701]  \n[[pdf]](static_detection\u002Fssd\u002FDSSD%20Deconvolutional%20Single%20Shot%20Detector%20ax1701.06659.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FDSSD.pdf)\n\n\u003Ca id=\"retinanet_\">\u003C\u002Fa>\n### RetinaNet\n- **Feature Pyramid Networks for Object Detection**  \n[ax1704]  \n[[pdf]](static_detection\u002Fretinanet\u002FFeature%20Pyramid%20Networks%20for%20Object%20Detection%20ax170419.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FFPN.pdf)\n- **Focal Loss for Dense Object Detection**  \n[ax180207\u002Ficcv17]  \n[[pdf]](static_detection\u002Fretinanet\u002FFocal%20Loss%20for%20Dense%20Object%20Detection%20ax180207%20iccv17.pdf)  \n[[notes]](static_detection\u002Fnotes\u002Ffocal_loss.pdf) \n\n\u003Ca id=\"anchor_free_\">\u003C\u002Fa>\n\n### 无锚框（Anchor Free）\n\n- **FoveaBox: 超越基于锚框的目标检测器（Beyond Anchor-based Object Detector）**  \n[ax1904]  \n[[pdf]](static_detection\u002Fanchor_free\u002FFoveaBox%20Beyond%20Anchor-based%20Object%20Detector%20ax1904.03797.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FFoveaBox%20Beyond%20Anchor-based%20Object%20Detector%20ax1904.03797.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Ftaokong\u002FFoveaBox)\n\n- **CornerNet: 将目标检测视为成对关键点（Detecting Objects as Paired Keypoints）**  \n[ax1903\u002Fijcv19]  \n[[pdf]](static_detection\u002Fanchor_free\u002FCornerNet%20Detecting%20Objects%20as%20Paired%20Keypoints%20ax1903%20ijcv19.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FCornerNet%20Detecting%20Objects%20as%20Paired%20Keypoints%20ax1903%20ijcv19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fprinceton-vl\u002FCornerNet)\n\n- **FCOS: 全卷积单阶段目标检测（Fully Convolutional One-Stage Object Detection）**  \n[ax1908\u002Ficcv19]  \n[[pdf]](static_detection\u002Fanchor_free\u002FFCOS%20Fully%20Convolutional%20One-Stage%20Object%20Detection%20ax1908%20iccv19.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FFCOS%20Fully%20Convolutional%20One-Stage%20Object%20Detection%20ax1908%20iccv19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Ftianzhi0549\u002FFCOS)  \n[[code\u002FFCOS_PLUS]](https:\u002F\u002Fgithub.com\u002Fyqyao\u002FFCOS_PLUS)  \n[[code\u002FVoVNet]](https:\u002F\u002Fgithub.com\u002Fvov-net\u002FVoVNet-FCOS)  \n[[code\u002FHRNet]](https:\u002F\u002Fgithub.com\u002FHRNet\u002FHRNet-FCOS)  \n[[code\u002FNAS]](https:\u002F\u002Fgithub.com\u002FLausannen\u002FNAS-FCOS)\n\n- **用于单阶段目标检测的特征选择无锚框模块（Feature Selective Anchor-Free Module for Single-Shot Object Detection）**  \n[ax1903\u002Fcvpr19]  \n[[pdf]](static_detection\u002Fanchor_free\u002FFeature%20Selective%20Anchor-Free%20Module%20for%20Single-Shot%20Object%20Detection%20ax1903.00621%20cvpr19.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FFeature%20Selective%20Anchor-Free%20Module%20for%20Single-Shot%20Object%20Detection%20ax1903.00621%20cvpr19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fhdjang\u002FFeature-Selective-Anchor-Free-Module-for-Single-Shot-Object-Detection)\n\n- **通过聚合极值点与中心点实现自底向上的目标检测（Bottom-up object detection by grouping extreme and center points）**  \n[ax1901]  \n[[pdf]](static_detection\u002Fanchor_free\u002FBottom-up%20object%20detection%20by%20grouping%20extreme%20and%20center%20points%201901.08043.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FBottom-up%20object%20detection%20by%20grouping%20extreme%20and%20center%20points%201901.08043.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FExtremeNet)\n\n- **通过自适应训练样本选择弥合基于锚框与无锚框检测之间的差距（Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection）**  \n[ax1912\u002Fcvpr20]  \n[[pdf]](static_detection\u002Fanchor_free\u002FBridging%20the%20Gap%20Between%20Anchor-based%20and%20Anchor-free%20Detection%20via%20Adaptive%20Training%20Sample%20Selection%201912.02424%20cvpr20.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FBridging%20the%20Gap%20Between%20Anchor-based%20and%20Anchor-free%20Detection%20via%20Adaptive%20Training%20Sample%20Selection%201912.02424%20cvpr20.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fsfzhang15\u002FATSS)\n\n- **使用 Transformer 实现端到端目标检测（End-to-end object detection with Transformers）**  \n[ax200528]  \n[[pdf]](static_detection\u002Fanchor_free\u002FEnd-to-End%20Object%20Detection%20with%20Transformers%20ax200528.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FEnd-to-end%20object%20detection%20with%20Transformers%20ax200528.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetr)\n\n- **将目标视为点（Objects as Points）**  \n[ax1904]  \n[[pdf]](static_detection\u002Fanchor_free\u002FObjects%20as%20Points%20ax1904.07850.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FObjects%20as%20Points%20ax1904.07850.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterNet)\n\n- **RepPoints：用于目标检测的点集表示（Point Set Representation for Object Detection）**  \n[iccv19]  \n[[pdf]](static_detection\u002Fanchor_free\u002FRepPoints%20Point%20Set%20Representation%20for%20Object%20Detection%201904.11490%20iccv19.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FRepPoints%20Point%20Set%20Representation%20for%20Object%20Detection%201904.11490%20iccv19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FRepPoints)\n\n\u003Ca id=\"mis_c_\">\u003C\u002Fa>\n### 其他（Misc）\n\n- **OverFeat：使用卷积网络实现集成的识别、定位与检测（Integrated Recognition, Localization and Detection using Convolutional Networks）**  \n[ax1402\u002Ficlr14]  \n[[pdf]](static_detection\u002FOverFeat%20Integrated%20Recognition,%20Localization%20and%20Detection%20using%20Convolutional%20Networks%20ax1402%20iclr14.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FOverFeat%20Integrated%20Recognition,%20Localization%20and%20Detection%20using%20Convolutional%20Networks%20ax1402%20iclr14.pdf)\n\n- **LSDA：通过迁移实现大规模检测（Large scale detection through adaptation）**  \n[ax1411\u002Fnips14]  \n[[pdf]](static_detection\u002FLSDA%20Large%20scale%20detection%20through%20adaptation%20nips14%20ax14_11.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FLSDA%20Large%20scale%20detection%20through%20adaptation%20nips14%20ax14_11.pdf)\n\n- **获取定位置信度以实现精确目标检测（Acquisition of Localization Confidence for Accurate Object Detection）**  \n[ax1807\u002Feccv18]  \n[[pdf]](static_detection\u002FAcquisition%20of%20Localization%20Confidence%20for%20Accurate%20Object%20Detection%201807.11590%20eccv18.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FIOU-Net.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fvacancy\u002FPreciseRoIPooling)\n\n- **EfficientDet：可扩展且高效的目标检测（Scalable and Efficient Object Detection）**  \n[cvpr20]  \n[[pdf]](static_detection\u002FEfficientDet_Scalable%20and%20efficient%20object%20detection.pdf)\n\n- **广义交并比（Generalized Intersection over Union）：一种用于边界框回归的度量与损失函数（A Metric and A Loss for Bounding Box Regression）**  \n[ax1902\u002Fcvpr19]  \n[[pdf]](static_detection\u002FGeneralized%20Intersection%20over%20Union%20A%20Metric%20and%20A%20Loss%20for%20Bounding%20Box%20Regression%201902.09630%20cvpr19.pdf)  \n[[notes]](static_detection\u002Fnotes\u002FGeneralized%20Intersection%20over%20Union%20A%20Metric%20and%20A%20Loss%20for%20Bounding%20Box%20Regression%201902.09630%20cvpr19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fgeneralized-iou)  \n[[project]](https:\u002F\u002Fgiou.stanford.edu\u002F)\n\n\u003Ca id=\"video_detectio_n_\">\u003C\u002Fa>\n## 视频目标检测（Video Detection）\n\n\u003Ca id=\"tubelet_\">\u003C\u002Fa>\n### Tubelet\n\n* **使用卷积神经网络从视频 Tubelet 中进行目标检测（Object Detection from Video Tubelets with Convolutional Neural Networks）**  \n[cvpr16]  \n[[pdf]](video_detection\u002Ftubelets\u002FObject_Detection_from_Video_Tubelets_with_Convolutional_Neural_Networks_CVPR16.pdf)  \n[[notes]](video_detection\u002Fnotes\u002FObject_Detection_from_Video_Tubelets_with_Convolutional_Neural_Networks_CVPR16.pdf)\n\n* **使用 Tubelet 提案网络进行视频目标检测（Object Detection in Videos with Tubelet Proposal Networks）**  \n[ax1704\u002Fcvpr17]  \n[[pdf]](video_detection\u002Ftubelets\u002FObject_Detection_in_Videos_with_Tubelet_Proposal_Networks_ax1704_cvpr17.pdf)  \n[[notes]](video_detection\u002Fnotes\u002FObject_Detection_in_Videos_with_Tubelet_Proposal_Networks_ax1704_cvpr17.pdf)\n\n\u003Ca id=\"fgf_a_\">\u003C\u002Fa>\n\n### FGFA\n* **Deep Feature Flow for Video Recognition**（用于视频识别的深度特征流）  \n[cvpr17]  \n[Microsoft Research]  \n[[pdf]](video_detection\u002Ffgfa\u002FDeep%20Feature%20Flow%20For%20Video%20Recognition%20cvpr17.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1611.07715)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FDeep-Feature-Flow)   \n* **Flow-Guided Feature Aggregation for Video Object Detection**（用于视频目标检测的光流引导特征聚合）  \n[ax1708\u002Ficcv17]  \n[[pdf]](video_detection\u002Ffgfa\u002FFlow-Guided%20Feature%20Aggregation%20for%20Video%20Object%20Detection%20ax1708%20iccv17.pdf)  \n[[notes]](video_detection\u002Fnotes\u002FFlow-Guided%20Feature%20Aggregation%20for%20Video%20Object%20Detection%20ax1708%20iccv17.pdf)  \n* **Towards High Performance Video Object Detection**（迈向高性能视频目标检测）  \n[ax1711]  \n[Microsoft]  \n[[pdf]](video_detection\u002Ffgfa\u002FTowards%20High%20Performance%20Video%20Object%20Detection%20ax171130%20microsoft.pdf)  \n[[notes]](video_detection\u002Fnotes\u002FTowards%20High%20Performance%20Video%20Object%20Detection%20ax171130%20microsoft.pdf)\n\n\u003Ca id=\"rnn_\">\u003C\u002Fa>\n### RNN\n* **Online Video Object Detection using Association LSTM**（使用关联 LSTM 的在线视频目标检测）  \n[iccv17]  \n[[pdf]](video_detection\u002Frnn\u002FOnline%20Video%20Object%20Detection%20using%20Association%20LSTM%20iccv17.pdf)  \n[[notes]](video_detection\u002Fnotes\u002FOnline%20Video%20Object%20Detection%20using%20Association%20LSTM%20iccv17.pdf)  \n* **Context Matters Refining Object Detection in Video with Recurrent Neural Networks**（上下文至关重要：利用循环神经网络优化视频中的目标检测）  \n[bmvc16]  \n[[pdf]](video_detection\u002Frnn\u002FContext%20Matters%20Refining%20Object%20Detection%20in%20Video%20with%20Recurrent%20Neural%20Networks%20bmvc16.pdf)  \n[[notes]](video_detection\u002Fnotes\u002FContext%20Matters%20Refining%20Object%20Detection%20in%20Video%20with%20Recurrent%20Neural%20Networks%20bmvc16.pdf)\n\n\u003Ca id=\"multi_object_tracking_\">\u003C\u002Fa>\n## 多目标跟踪（Multi Object Tracking）\n\n\u003Ca id=\"joint_detection_\">\u003C\u002Fa>\n### 联合检测（Joint-Detection）\n\n* **Tracking Objects as Points**（将目标作为点进行跟踪）  \n[ax2004]  \n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FTracking%20Objects%20as%20Points%202004.01177.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking%20Objects%20as%20Points%202004.01177.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterTrack)[pytorch]\n\n\n\u003Ca id=\"identity_embeddin_g_\">\u003C\u002Fa>\n#### 身份嵌入（Identity Embedding）\n\n* **MOTS Multi-Object Tracking and Segmentation**（MOTS：多目标跟踪与分割）  \n[cvpr19]  \n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FMOTS%20Multi-Object%20Tracking%20and%20Segmentation%20ax1904%20cvpr19.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FMOTS%20Multi-Object%20Tracking%20and%20Segmentation%20ax1904%20cvpr19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FTrackR-CNN)  \n[[project\u002Fdata]](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fmots)  \n* **Towards Real-Time Multi-Object Tracking**（迈向实时多目标跟踪）  \n[ax1909]  \n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FTowards%20Real-Time%20Multi-Object%20Tracking%20ax1909.12605v1.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FTowards%20Real-Time%20Multi-Object%20Tracking%20ax1909.12605v1.pdf)  \n* **A Simple Baseline for Multi-Object Tracking**（多目标跟踪的一个简单基线）  \n[ax2004]  \n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FA%20Simple%20Baseline%20for%20Multi-Object%20Tracking%202004.01888.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FA%20Simple%20Baseline%20for%20Multi-Object%20Tracking%202004.01888.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fifzhang\u002FFairMOT)  \n\n* **Integrated Object Detection and Tracking with Tracklet-Conditioned Detection**（基于轨迹片段条件检测的集成目标检测与跟踪）  \n[ax1811]  \n[[pdf]](multi_object_tracking\u002Fjoint_detection\u002FIntegrated%20Object%20Detection%20and%20Tracking%20with%20Tracklet-Conditioned%20Detection%201811.11167.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FIntegrated%20Object%20Detection%20and%20Tracking%20with%20Tracklet-Conditioned%20Detection%201811.11167.pdf)\n\n\n\n\u003Ca id=\"association_\">\u003C\u002Fa>\n### 关联（Association）\n\n* **Deep Affinity Network for Multiple Object Tracking**（用于多目标跟踪的深度亲和网络）  \n[ax1810\u002Ftpami19]  \n[[pdf]](multi_object_tracking\u002Fassociation\u002FDeep%20Affinity%20Network%20for%20Multiple%20Object%20Tracking%20ax1810.11780%20tpami19.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FDeep%20Affinity%20Network%20for%20Multiple%20Object%20Tracking%20ax1810.11780%20tpami19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002FshijieS\u002FSST) [pytorch]\n\n\u003Ca id=\"deep_learning_\">\u003C\u002Fa>\n### 深度学习（Deep Learning）\n\n* **Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal Attention Mechanism**（使用基于 CNN 的单目标跟踪器与时空注意力机制的在线多目标跟踪）  \n[ax1708\u002Ficcv17]  \n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FOnline%20Multi-Object%20Tracking%20Using%20CNN-based%20Single%20Object%20Tracker%20with%20Spatial-Temporal%20Attention%20Mechanism%201708.02843%20iccv17.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1708.02843)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FOnline%20Multi-Object%20Tracking%20Using%20CNN-based%20Single%20Object%20Tracker%20with%20Spatial-Temporal%20Attention%20Mechanism%201708.02843%20iccv17.pdf)  \n* **Online multi-object tracking with dual matching attention networks**（使用双匹配注意力网络的在线多目标跟踪）  \n[ax1902\u002Feccv18]  \n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FOnline%20multi-object%20tracking%20with%20dual%20matching%20attention%20networks%201902.00749%20eccv18.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1902.00749)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FOnline%20multi-object%20tracking%20with%20dual%20matching%20attention%20networks%201902.00749%20eccv18.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fjizhu1023\u002FDMAN_MOT)  \n* **FAMNet Joint Learning of Feature, Affinity and Multi-Dimensional Assignment for Online Multiple Object Tracking**（FAMNet：用于在线多目标跟踪的特征、亲和性与多维分配联合学习）  \n[iccv19]  \n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FFAMNet%20Joint%20Learning%20of%20Feature,%20Affinity%20and%20Multi-Dimensional%20Assignment%20for%20Online%20Multiple%20Object%20Tracking%20iccv19.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FFAMNet%20Joint%20Learning%20of%20Feature,%20Affinity%20and%20Multi-Dimensional%20Assignment%20for%20Online%20Multiple%20Object%20Tracking%20iccv19.pdf)  \n\n* **Exploit the Connectivity: Multi-Object Tracking with TrackletNet**（利用连通性：使用 TrackletNet 进行多目标跟踪）  \n[ax1811\u002Fmm19]  \n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FExploit%20the%20Connectivity%20Multi-Object%20Tracking%20with%20TrackletNet%20ax1811.07258%20mm19.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FExploit%20the%20Connectivity%20Multi-Object%20Tracking%20with%20TrackletNet%20ax1811.07258%20mm19.pdf)  \n* **Tracking without bells and whistles**（简洁高效的跟踪方法）  \n[ax1903\u002Ficcv19]  \n[[pdf]](multi_object_tracking\u002Fdeep_learning\u002FTracking%20without%20bells%20and%20whistles%20ax1903.05625%20iccv19.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking%20without%20bells%20and%20whistles%20ax1903.05625%20iccv19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fphil-bergmann\u002Ftracking_wo_bnw) [pytorch]\n\n\u003Ca id=\"rnn__1\">\u003C\u002Fa>\n\n### RNN（循环神经网络）\n* **Tracking The Untrackable: Learning To Track Multiple Cues with Long-Term Dependencies**  \n[ax1704\u002Ficcv17]  \n[Stanford]  \n[[pdf]](multi_object_tracking\u002Frnn\u002FTracking%20The%20Untrackable%20Learning%20To%20Track%20Multiple%20Cues%20with%20Long-Term%20Dependencies%20ax17_4_iccv17.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking_The_Untrackable_Learning_To_Track_Multiple_Cues_with_Long-Term_Dependencies.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1701.01909)  \n[[project]](http:\u002F\u002Fweb.stanford.edu\u002F~alahi\u002F)  \n* **Multi-object Tracking with Neural Gating Using Bilinear LSTM**  \n[eccv18]  \n[[pdf]](multi_object_tracking\u002Frnn\u002FMulti-object%20Tracking%20with%20Neural%20Gating%20Using%20Bilinear%20LSTM_eccv18.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FMulti-object%20Tracking%20with%20Neural%20Gating%20Using%20Bilinear%20LSTM_eccv18.pdf)  \n* **Eliminating Exposure Bias and Metric Mismatch in Multiple Object Tracking**  \n[cvpr19]  \n[[pdf]](multi_object_tracking\u002Frnn\u002FEliminating%20Exposure%20Bias%20and%20Metric%20Mismatch%20in%20Multiple%20Object%20Tracking%20cvpr19.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FEliminating%20Exposure%20Bias%20and%20Metric%20Mismatch%20in%20Multiple%20Object%20Tracking%20cvpr19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fmaksay\u002Fseq-train)\n\n\u003Ca id=\"unsupervised_learning_\">\u003C\u002Fa>\n### 无监督学习（Unsupervised Learning）\n* **Unsupervised Person Re-identification by Deep Learning Tracklet Association**  \n[ax1809\u002Feccv18]  \n[[pdf]](multi_object_tracking\u002Funsupervised\u002FUnsupervised%20Person%20Re-identification%20by%20Deep%20Learning%20Tracklet%20Association%201809.02874%20eccv18.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FUnsupervised%20Person%20Re-identification%20by%20Deep%20Learning%20Tracklet%20Association%201809.02874%20eccv18.pdf)  \n* **Tracking by Animation: Unsupervised Learning of Multi-Object Attentive Trackers**  \n[ax1809\u002Fcvpr19]  \n[[pdf]](multi_object_tracking\u002Funsupervised\u002FTracking%20by%20Animation%20Unsupervised%20Learning%20of%20Multi-Object%20Attentive%20Trackers%20cvpr19%20ax1809.03137.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1809.03137)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FTracking%20by%20Animation%20Unsupervised%20Learning%20of%20Multi-Object%20Attentive%20Trackers%20cvpr19%20ax1809.03137.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fzhen-he\u002Ftracking-by-animation)  \n* **Simple Unsupervised Multi-Object Tracking**  \n[ax2006]  \n[[pdf]](multi_object_tracking\u002Funsupervised\u002FSimple%20Unsupervised%20Multi-Object%20Tracking%202006.02609.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FSimple%20Unsupervised%20Multi-Object%20Tracking%202006.02609.pdf)\n\n\u003Ca id=\"reinforcement_learning_\">\u003C\u002Fa>\n### 强化学习（Reinforcement Learning）\n* **Learning to Track: Online Multi-object Tracking by Decision Making**  \n[iccv15]  \n[Stanford]  \n[[pdf]](multi_object_tracking\u002Frl\u002FLearning%20to%20Track%20Online%20Multi-object%20Tracking%20by%20Decision%20Making%20%20iccv15.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FLearning_to_Track_Online_Multi-object_Tracking_by_Decision_Making__iccv15.pdf)  \n[[code (matlab)]](https:\u002F\u002Fgithub.com\u002Fyuxng\u002FMDP_Tracking)  \n[[project]](https:\u002F\u002Fyuxng.github.io\u002F)  \n* **Collaborative Deep Reinforcement Learning for Multi-Object Tracking**  \n[eccv18]  \n[[pdf]](multi_object_tracking\u002Frl\u002FCollaborative%20Deep%20Reinforcement%20Learning%20for%20Multi-Object%20Tracking_eccv18.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FCollaborative%20Deep%20Reinforcement%20Learning%20for%20Multi-Object%20Tracking_eccv18.pdf)\n\n\u003Ca id=\"network_flow_\">\u003C\u002Fa>\n### 网络流（Network Flow）\n* **Near-Online Multi-target Tracking with Aggregated Local Flow Descriptor**  \n[iccv15]  \n[NEC Labs]  \n[[pdf]](multi_object_tracking\u002Fnetwork_flow\u002FNear-online%20multi-target%20tracking%20with%20aggregated%20local%20%EF%AC%82ow%20descriptor%20iccv15.pdf)  \n[[author]](http:\u002F\u002Fwww-personal.umich.edu\u002F~wgchoi\u002F)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FNOMT.pdf)  \n* **Deep Network Flow for Multi-Object Tracking**  \n[cvpr17]  \n[NEC Labs]  \n[[pdf]](multi_object_tracking\u002Fnetwork_flow\u002FDeep%20Network%20Flow%20for%20Multi-Object%20Tracking%20cvpr17.pdf)  \n[[supplementary]](multi_object_tracking\u002Fnetwork_flow\u002FDeep%20Network%20Flow%20for%20Multi-Object%20Tracking%20cvpr17_supplemental.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FDeep%20Network%20Flow%20for%20Multi-Object%20Tracking%20cvpr17.pdf)  \n* **Learning a Neural Solver for Multiple Object Tracking**  \n[ax1912\u002Fcvpr20]  \n[[pdf]](multi_object_tracking\u002Fnetwork_flow\u002FLearning%20a%20Neural%20Solver%20for%20Multiple%20Object%20Tracking%201912.07515%20cvpr20.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FLearning%20a%20Neural%20Solver%20for%20Multiple%20Object%20Tracking%201912.07515%20cvpr20.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fdvl-tum\u002Fmot_neural_solver)\n\n\u003Ca id=\"graph_optimization_\">\u003C\u002Fa>\n### 图优化（Graph Optimization）\n* **A Multi-cut Formulation for Joint Segmentation and Tracking of Multiple Objects**  \n[ax1607]  \n[highest MT on MOT2015]  \n[University of Freiburg, Germany]  \n[[pdf]](multi_object_tracking\u002Fbatch\u002FA%20Multi-cut%20Formulation%20for%20Joint%20Segmentation%20and%20Tracking%20of%20Multiple%20Objects%20ax16_9%20%5Bbest%20MT%20on%20MOT15%5D.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1607.06317)  \n[[author]](https:\u002F\u002Flmb.informatik.uni-freiburg.de\u002Fpeople\u002Fkeuper\u002Fpublications.html)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FA_Multi-cut_Formulation_for_Joint_Segmentation_and_Tracking_of_Multiple_Objects.pdf)\n\n\u003Ca id=\"baselin_e_\">\u003C\u002Fa>\n### 基线方法（Baseline）\n* **Simple Online and Realtime Tracking**  \n[icip16]  \n[[pdf]](multi_object_tracking\u002Fbaseline\u002FSimple%20Online%20and%20Realtime%20Tracking%20ax1707%20icip16.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FSimple%20Online%20and%20Realtime%20Tracking%20ax1707%20icip16.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fabewley\u002Fsort)  \n* **High-Speed Tracking-by-Detection Without Using Image Information**  \n[avss17]  \n[[pdf]](multi_object_tracking\u002Fbaseline\u002FHigh-Speed%20Tracking-by-Detection%20Without%20Using%20Image%20Information%20avss17.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FHigh-Speed%20Tracking-by-Detection%20Without%20Using%20Image%20Information%20avss17.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fbochinski\u002Fiou-tracker)\n     \n\u003Ca id=\"metrics_\">\u003C\u002Fa>\n### 评估指标（Metrics）\n* **HOTA: A Higher Order Metric for Evaluating Multi-object Tracking**  \n[ijcv20\u002F08]  \n[[pdf]](multi_object_tracking\u002Fmetrics\u002FHOTA%20A%20Higher%20Order%20Metric%20for%20Evaluating%20Multi-object%20Tracking%20sl_open_2010_ijcv2008.pdf)  \n[[notes]](multi_object_tracking\u002Fnotes\u002FHOTA%20A%20Higher%20Order%20Metric%20for%20Evaluating%20Multi-object%20Tracking%20sl_open_2010_ijcv2008.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002FJonathonLuiten\u002FHOTA-metrics)\n\n\u003Ca id=\"single_object_tracking_\">\u003C\u002Fa>\n## 单目标跟踪（Single Object Tracking）\n\n\u003Ca id=\"reinforcement_learning__1\">\u003C\u002Fa>\n\n### 强化学习（Reinforcement Learning）\n* **Deep Reinforcement Learning for Visual Object Tracking in Videos**  \n[ax1704] [加州大学圣塔芭芭拉分校（USC-Santa Barbara），三星研究院（Samsung Research）]  \n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FDeep%20Reinforcement%20Learning%20for%20Visual%20Object%20Tracking%20in%20Videos%20ax17_4.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1701.08936)  \n[[author]](http:\u002F\u002Fwww.cs.ucsb.edu\u002F~dazhang\u002F)  \n[[notes]](single_object_tracking\u002Fnotes\u002FDeep_Reinforcement_Learning_for_Visual_Object_Tracking_in_Videos.pdf)  \n* **Visual Tracking by Reinforced Decision Making**  \n[ax1702] [首尔国立大学（Seoul National University），中央大学（Chung-Ang University）]  \n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FVisual%20Tracking%20by%20Reinforced%20Decision%20Making%20ax17_2.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1702.06291)  \n[[author]](http:\u002F\u002Fcau.ac.kr\u002F~jskwon\u002F)  \n[[notes]](single_object_tracking\u002Fnotes\u002FVisual_Tracking_by_Reinforced_Decision_Making_ax17.pdf)  \n* **Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning**  \n[cvpr17] [首尔国立大学（Seoul National University）]  \n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FAction-Decision%20Networks%20for%20Visual%20Tracking%20with%20Deep%20Reinforcement%20Learning%20%20cvpr17%20supplementary.pdf)  \n[[supplementary]](single_object_tracking\u002Freinforcement_learning\u002FAction-Decision%20Networks%20for%20Visual%20Tracking%20with%20Deep%20Reinforcement%20Learning%20%20cvpr17.pdf)  \n[[project]](https:\u002F\u002Fsites.google.com\u002Fview\u002Fcvpr2017-adnet)  \n[[notes]](single_object_tracking\u002Fnotes\u002FAction-Decision_Networks_for_Visual_Tracking_with_Deep_Reinforcement_Learning_cvpr17.pdf)   \n[[code]](https:\u002F\u002Fgithub.com\u002Fildoonet\u002Ftf-adnet-tracking)   \n* **End-to-end Active Object Tracking via Reinforcement Learning**  \n[ax1705]  \n[北京大学（Peking University），腾讯 AI Lab（Tencent AI Lab）]  \n[[pdf]](single_object_tracking\u002Freinforcement_learning\u002FEnd-to-end%20Active%20Object%20Tracking%20via%20Reinforcement%20Learning%20ax17_5.pdf)  \n[[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1705.10561)\n\n\u003Ca id=\"siamese_\">\u003C\u002Fa>\n### Siamese 网络\n* **Fully-Convolutional Siamese Networks for Object Tracking**  \n[eccv16]  \n[[pdf]](single_object_tracking\u002Fsiamese\u002FFully-Convolutional%20Siamese%20Networks%20for%20Object%20Tracking%20eccv16_9.pdf)  \n[[project]](https:\u002F\u002Fwww.robots.ox.ac.uk\u002F~luca\u002Fsiamese-fc.html)  \n[[notes]](single_object_tracking\u002Fnotes\u002FSiameseFC.pdf)  \n* **High Performance Visual Tracking with Siamese Region Proposal Network**  \n[cvpr18]  \n[[pdf]](single_object_tracking\u002Fsiamese\u002FHigh%20Performance%20Visual%20Tracking%20with%20Siamese%20Region%20Proposal%20Network_cvpr18.pdf)  \n[[author]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~qwang\u002F)  \n[[notes]](single_object_tracking\u002Fnotes\u002FHigh%20Performance%20Visual%20Tracking%20with%20Siamese%20Region%20Proposal%20Network_cvpr18.pdf)  \n* **Siam R-CNN Visual Tracking by Re-Detection**  \n[cvpr20]  \n[[pdf]](single_object_tracking\u002Fsiamese\u002FSiam%20R-CNN%20Visual%20Tracking%20by%20Re-Detection%201911.12836%20cvpr20.pdf)  \n[[notes]](single_object_tracking\u002Fnotes\u002FSiam%20R-CNN%20Visual%20Tracking%20by%20Re-Detection%201911.12836%20cvpr20.pdf)  \n[[project]](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fsiamrcnn)  \n[[code]](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FSiamR-CNN)  \n\n\u003Ca id=\"correlation_\">\u003C\u002Fa>\n### 相关滤波（Correlation）\n* **ATOM: Accurate Tracking by Overlap Maximization**  \n[cvpr19]  \n[[pdf]](single_object_tracking\u002Fcorrelation\u002FATOM%20Accurate%20Tracking%20by%20Overlap%20Maximization%20ax1811.07628%20cvpr19.pdf)  \n[[notes]](single_object_tracking\u002Fnotes\u002FATOM%20Accurate%20Tracking%20by%20Overlap%20Maximization%20ax1811.07628%20cvpr19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking)  \n* **DiMP: Learning Discriminative Model Prediction for Tracking**  \n[iccv19]  \n[[pdf]](single_object_tracking\u002Fcorrelation\u002FDiMP%20Learning%20Discriminative%20Model%20Prediction%20for%20Tracking%20ax1904.07220%20iccv19.pdf)  \n[[notes]](single_object_tracking\u002Fnotes\u002FDiMP%20Learning%20Discriminative%20Model%20Prediction%20for%20Tracking%20ax1904.07220%20iccv19.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking)  \n* **D3S – A Discriminative Single Shot Segmentation Tracker**  \n[cvpr20]  \n[[pdf]](single_object_tracking\u002Fcorrelation\u002FD3S%20–%20A%20Discriminative%20Single%20Shot%20Segmentation%20Tracker%201911.08862v1%20cvpr20.pdf)  \n[[notes]](single_object_tracking\u002Fnotes\u002FD3S%20–%20A%20Discriminative%20Single%20Shot%20Segmentation%20Tracker%201911.08862v1%20cvpr20.pdf)  \n[[code]](https:\u002F\u002Fgithub.com\u002Falanlukezic\u002Fd3s)\n\n\u003Ca id=\"mis_c__1\">\u003C\u002Fa>\n### 其他（Misc）\n\n* **Bridging the Gap Between Detection and Tracking: A Unified Approach**  \n[iccv19]  \n[[pdf]](single_object_tracking\u002FBridging%20the%20Gap%20Between%20Detection%20and%20Tracking%20A%20Unified%20Approach%20iccv19.pdf)  \n[[notes]](single_object_tracking\u002Fnotes\u002FBridging%20the%20Gap%20Between%20Detection%20and%20Tracking%20A%20Unified%20Approach%20iccv19.pdf)\n\n\u003Ca id=\"deep_learning__1\">\u003C\u002Fa>\n## 深度学习（Deep Learning）\n\n- **Do Deep Nets Really Need to be Deep**  \n[nips14]  \n[[pdf]](deep_learning\u002Ftheory\u002FDo%20Deep%20Nets%20Really%20Need%20to%20be%20Deep%20ax1410%20nips14.pdf)  \n[[notes]](deep_learning\u002Fnotes\u002FDo%20Deep%20Nets%20Really%20Need%20to%20be%20Deep%20ax1410%20nips14.pdf)\n\n\u003Ca id=\"synthetic_gradient_s_\">\u003C\u002Fa>\n### 合成梯度（Synthetic Gradients）\n- **Decoupled Neural Interfaces using Synthetic Gradients**  \n[ax1608]  \n[[pdf]](deep_learning\u002Fsynthetic_gradients\u002FDecoupled%20Neural%20Interfaces%20using%20Synthetic%20Gradients%20ax1608.05343.pdf)  \n[[notes]](deep_learning\u002Fnotes\u002FDecoupled%20Neural%20Interfaces%20using%20Synthetic%20Gradients%20ax1608.05343.pdf)    \n- **Understanding Synthetic Gradients and Decoupled Neural Interfaces**  \n[ax1703]  \n[[pdf]](deep_learning\u002Fsynthetic_gradients\u002FUnderstanding%20Synthetic%20Gradients%20and%20Decoupled%20Neural%20Interfaces%20ax1703.00522.pdf)  \n[[notes]](deep_learning\u002Fnotes\u002FUnderstanding%20Synthetic%20Gradients%20and%20Decoupled%20Neural%20Interfaces%20ax1703.00522.pdf)\n\n\u003Ca id=\"efficient_\">\u003C\u002Fa>\n### 高效模型（Efficient）\n- **EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks**  \n[icml2019]  \n[[pdf]](deep_learning\u002Fefficient\u002FEfficientNet_Rethinking%20model%20scaling%20for%20CNNs.pdf)  \n[[notes]](deep_learning\u002Fnotes\u002FEfficientNet_%20Rethinking%20Model%20Scaling%20for%20Convolutional%20Neural%20Networks.pdf)\n\n\u003Ca id=\"unsupervised_learning__1\">\u003C\u002Fa>\n## 无监督学习（Unsupervised Learning）\n- **Learning Features by Watching Objects Move**  \n(cvpr17)  \n[[pdf]](unsupervised\u002Fsegmentation\u002FLearning%20Features%20by%20Watching%20Objects%20Move%20ax170412%20cvpr17.pdf)  \n[[notes]](unsupervised\u002Fnotes\u002FLearning%20Features%20by%20Watching%20Objects%20Move%20ax170412%20cvpr17.pdf)\n    \n\u003Ca id=\"interpolation_\">\u003C\u002Fa>\n## 插值（Interpolation）\n- **Video Frame Interpolation via Adaptive Convolution**  \n[cvpr17 \u002F iccv17]  \n[[pdf (cvpr17)]](interpolation\u002FVideo%20Frame%20Interpolation%20via%20Adaptive%20Convolution%20ax1703.pdf)  \n[[pdf (iccv17)]](interpolation\u002FVideo%20Frame%20Interpolation%20via%20Adaptive%20Separable%20Convolution%20iccv17.pdf)  \n[[ppt]](interpolation\u002Fnotes\u002FVideo%20Frame%20Interpolation%20via%20Adaptive%20Convolution%20ax1703.pdf)\n\n\u003Ca id=\"autoencoder_\">\u003C\u002Fa>\n## 自编码器（Autoencoder）\n\n\u003Ca id=\"variational_\">\u003C\u002Fa>\n\n### 变分方法（Variational）\n- **beta-VAE：使用约束变分框架学习基本视觉概念** [iclr17]  \n[[pdf]](autoencoder\u002Fvariational\u002Fbeta-VAE%20Learning%20Basic%20Visual%20Concepts%20with%20a%20Constrained%20Variational%20Framework%20iclr17.pdf)  \n[[notes]](autoencoder\u002Fnotes\u002Fbeta-VAE%20Learning%20Basic%20Visual%20Concepts%20with%20a%20Constrained%20Variational%20Framework%20iclr17.pdf)\n- **通过因子分解实现解耦（Disentangling by Factorising）** [ax1806]  \n[[pdf]](autoencoder\u002Fvariational\u002FDisentangling%20by%20Factorising%20ax1806.pdf)  \n[[notes]](autoencoder\u002Fnotes\u002FDisentangling%20by%20Factorising%20ax1806.pdf)  \n\n\u003Ca id=\"dataset_s_\">\u003C\u002Fa>\n# 数据集\n\n\u003Ca id=\"multi_object_tracking__1\">\u003C\u002Fa>\n## 多目标跟踪（Multi Object Tracking）\n\n- [IDOT](https:\u002F\u002Fgithub.com\u002Fbitslab\u002FIDOT_dataset)\n- [UA-DETRAC Benchmark Suite](http:\u002F\u002Fdetrac-db.rit.albany.edu\u002F)\n- [GRAM Road-Traffic Monitoring](http:\u002F\u002Fagamenon.tsc.uah.es\u002FPersonales\u002Frlopez\u002Fdata\u002Frtm\u002F) [[paper]](https:\u002F\u002Flink.springer.com\u002Fchapter\u002F10.1007\u002F978-3-642-38622-0_32)\n- [Ko-PER Intersection Dataset](http:\u002F\u002Fwww.uni-ulm.de\u002Fin\u002Fmrm\u002Fforschung\u002Fdatensaetze.html)\n- [TRANCOS](http:\u002F\u002Fagamenon.tsc.uah.es\u002FPersonales\u002Frlopez\u002Fdata\u002Ftrancos\u002F)\n- [Urban Tracker](https:\u002F\u002Fwww.jpjodoin.com\u002Furbantracker\u002Fdataset.html)\n- [DARPA VIVID \u002F PETS 2005](http:\u002F\u002Fvision.cse.psu.edu\u002Fdata\u002FvividEval\u002Fdatasets\u002Fdatasets.html) [非固定相机]\n- [KIT-AKS](http:\u002F\u002Fi21www.ira.uka.de\u002Fimage_sequences\u002F) [无真值标注]\n- [CBCL StreetScenes Challenge Framework](http:\u002F\u002Fcbcl.mit.edu\u002Fsoftware-datasets\u002Fstreetscenes\u002F) [无俯视视角]\n- [MOT 2015](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002F2D_MOT_2015\u002F) [主要为街面视角]\n- [MOT 2016](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FMOT16\u002F) [主要为街面视角]\n- [MOT 2017](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FMOT17\u002F) [主要为街面视角]\n- [MOT 2020](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FMOT20\u002F) [主要为俯视视角]\n- [MOTS: 多目标跟踪与分割（Multi-Object Tracking and Segmentation）](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fmots) [基于 MOT 和 KITTI]\n- [CVPR 2019](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002F11) [主要为街面视角]\n- [PETS 2009](http:\u002F\u002Fwww.cvg.reading.ac.uk\u002FPETS2009\u002Fa.html) [不含车辆]\n- [PETS 2017](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FPETS2017\u002F) [低密度] [主要为行人]\n- [DukeMTMC](http:\u002F\u002Fvision.cs.duke.edu\u002FDukeMTMC\u002F) [多相机] [静态背景] [行人] [高于街面视角] [网站无法访问]\n- [KITTI Tracking Dataset](http:\u002F\u002Fwww.cvlibs.net\u002Fdatasets\u002Fkitti\u002Feval_tracking.php) [无俯视视角] [非固定相机]\n- [The WILDTRACK Seven-Camera HD Dataset](https:\u002F\u002Fcvlab.epfl.ch\u002Fdata\u002Fdata-wildtrack\u002F) [行人检测与跟踪]\n- [基于移动平台的 3D 交通场景理解（3D Traffic Scene Understanding from Movable Platforms）](http:\u002F\u002Fwww.cvlibs.net\u002Fprojects\u002Fintersection\u002F) [交叉路口交通] [立体视觉设置] [移动相机]\n- [LOST : 场景长期观测与轨迹数据集（Longterm Observation of Scenes with Tracks）](http:\u002F\u002Flost.cse.wustl.edu\u002F) [俯视与街面视角] [无真值标注]\n- [JTA](http:\u002F\u002Fimagelab.ing.unimore.it\u002Fimagelab\u002Fpage.asp?IdPage=25) [俯视与街面视角] [合成数据\u002FGTA 5] [行人] [3D 标注]\n- [PathTrack: 使用路径监督进行快速轨迹标注（Fast Trajectory Annotation with Path Supervision）](http:\u002F\u002Fpeople.ee.ethz.ch\u002F~daid\u002Fpathtrack\u002F) [俯视与街面视角] [iccv17] [行人]\n- [CityFlow](https:\u002F\u002Fwww.aicitychallenge.org\u002F) [杆装摄像头] [交叉路口] [车辆] [重识别（re-id）] [cvpr19]\n- [JackRabbot Dataset](https:\u002F\u002Fjrdb.stanford.edu\u002F) [RGBD] [正面视角][室内\u002F室外][斯坦福]\n- [TAO: 面向任意目标跟踪的大规模基准（A Large-Scale Benchmark for Tracking Any Object）](http:\u002F\u002Ftaodataset.org\u002F) [eccv20] [[code]](https:\u002F\u002Fgithub.com\u002FTAO-Dataset\u002Ftao)\n- [爱丁堡办公室监控视频数据集（Edinburgh office monitoring video dataset）](http:\u002F\u002Fhomepages.inf.ed.ac.uk\u002Frbf\u002FOFFICEDATA\u002F) [室内][长期][人物大多静止]\n- [Waymo Open Dataset](https:\u002F\u002Fwaymo.com\u002Fopen\u002F) [室外][车辆]\n\n\u003Ca id=\"uav_\">\u003C\u002Fa>\n### 无人机（UAV）\n\n- [斯坦福无人机数据集（Stanford Drone Dataset）](http:\u002F\u002Fcvgl.stanford.edu\u002Fprojects\u002Fuav_data\u002F)\n- [UAVDT - 无人机基准：目标检测与跟踪（The Unmanned Aerial Vehicle Benchmark: Object Detection and Tracking）](https:\u002F\u002Fsites.google.com\u002Fsite\u002Fdaviddo0323\u002Fprojects\u002Fuavdt) [无人机] [交叉路口\u002F高速公路] [车辆] [eccv18]\n- [VisDrone](https:\u002F\u002Fgithub.com\u002FVisDrone\u002FVisDrone-Dataset)\n\n\u003Ca id=\"synthetic_\">\u003C\u002Fa>\n### 合成数据（Synthetic）\n\n- [MNIST-MOT \u002F MNIST-Sprites](https:\u002F\u002Fgithub.com\u002Fzhen-he\u002Ftracking-by-animation) [脚本生成] [cvpr19]\n- [TUB 多目标多相机跟踪数据集（TUB Multi-Object and Multi-Camera Tracking Dataset）](https:\u002F\u002Fwww.nue.tu-berlin.de\u002Fmenue\u002Fforschung\u002Fsoftware_und_datensaetze\u002Fmocat\u002F) [avss16]\n- [Virtual KITTI](http:\u002F\u002Fwww.xrce.xerox.com\u002FResearch-Development\u002FComputer-Vision\u002FProxy-Virtual-Worlds) [[arxiv]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1605.06457) [cvpr16] [链接似乎已失效]\n\n\u003Ca id=\"microscopy___cell_tracking_\">\u003C\u002Fa>\n### 显微镜 \u002F 细胞跟踪（Microscopy \u002F Cell Tracking）\n\n- [细胞跟踪挑战赛（Cell Tracking Challenge）](http:\u002F\u002Fcelltrackingchallenge.net\u002F) [nature methods\u002F2017]\n- [CTMC: 带有有丝分裂检测的细胞跟踪数据集挑战（Cell Tracking with Mitosis Detection Dataset Challenge）](https:\u002F\u002Fivc.ischool.utexas.edu\u002Fctmc\u002F) [cvprw20] [[MOT]](https:\u002F\u002Fmotchallenge.net\u002Fdata\u002FCTMC-v1\u002F)\n\n\u003Ca id=\"single_object_tracking__1\">\u003C\u002Fa>\n## 单目标跟踪（Single Object Tracking）\n\n- [TrackingNet: 面向真实场景的大规模目标跟踪数据集与基准](https:\u002F\u002Ftracking-net.org\u002F) [eccv18]\n- [LaSOT: 大规模单目标跟踪（Large-scale Single Object Tracking）](https:\u002F\u002Fcis.temple.edu\u002Flasot\u002F) [cvpr19]\n- [Need for speed: 高帧率目标跟踪基准（A benchmark for higher frame rate object tracking）](http:\u002F\u002Fci2cv.net\u002Fnfs\u002Findex.html) [iccv17]\n- [野外长期跟踪基准（Long-term Tracking in the Wild A Benchmark）](https:\u002F\u002Foxuva.github.io\u002Flong-term-tracking-benchmark\u002F) [eccv18]\n- [UAV123: 无人机跟踪基准与模拟器（A benchmark and simulator for UAV tracking）](https:\u002F\u002Fuav123.org\u002F) [eccv16] [[project]](https:\u002F\u002Fivul.kaust.edu.sa\u002FPages\u002Fpub-benchmark-simulator-uav.aspx)\n- [Sim4CV: 面向计算机视觉应用的照片级真实感模拟器（A Photo-Realistic Simulator for Computer Vision Applications）](https:\u002F\u002Fsim4cv.org\u002F) [ijcv18]\n- [CDTB: 彩色与深度视觉目标跟踪与基准（A Color and Depth Visual Object Tracking and Benchmark）](https:\u002F\u002Fwww.vicos.si\u002FProjects\u002FCDTB) [iccv19] [RGBD]\n- [Temple Color 128 - 彩色跟踪基准（Color Tracking Benchmark）](http:\u002F\u002Fwww.dabi.temple.edu\u002F~hbling\u002Fdata\u002FTColor-128\u002FTColor-128.html) [tip15]\n\n\u003Ca id=\"video_detectio_n__1\">\u003C\u002Fa>\n## 视频检测（Video Detection）\n\n- [YouTube-BB](https:\u002F\u002Fresearch.google.com\u002Fyoutube-bb\u002Fdownload.html)\n- [Imagenet-VID](http:\u002F\u002Fbvisionweb1.cs.unc.edu\u002Filsvrc2015\u002Fdownload-videos-3j16.php)\n\n\u003Ca id=\"video_understanding___activity_recognitio_n_\">\u003C\u002Fa>\n### 视频理解 \u002F 行为识别（Video Understanding \u002F Activity Recognition）\n\n- [YouTube-8M](https:\u002F\u002Fresearch.google.com\u002Fyoutube8m\u002F)\n- [AVA: 原子视觉动作视频数据集（A Video Dataset of Atomic Visual Action）](https:\u002F\u002Fresearch.google.com\u002Fava\u002F)\n- [VIRAT 视频数据集（VIRAT Video Dataset）](http:\u002F\u002Fwww.viratdata.org\u002F)\n- [Kinetics 行为识别数据集（Kinetics Action Recognition Dataset）](https:\u002F\u002Fdeepmind.com\u002Fresearch\u002Fopen-source\u002Fkinetics)\n\n\u003Ca id=\"static_detectio_n__1\">\u003C\u002Fa>\n\n## 静态检测（Static Detection）\n- [PASCAL Visual Object Classes](http:\u002F\u002Fhost.robots.ox.ac.uk\u002Fpascal\u002FVOC\u002F)\n- [A Large-Scale Dataset for Vehicle Re-Identification in the Wild](https:\u002F\u002Fgithub.com\u002FPKU-IMRE\u002FVERI-Wild)  \n  [cvpr19]\n- [Object Detection-based annotations for some frames of the VIRAT dataset](https:\u002F\u002Fgithub.com\u002Fahrnbom\u002FViratAnnotationObjectDetection)\n- [MIO-TCD: A new benchmark dataset for vehicle classification and localization](http:\u002F\u002Fpodoce.dinf.usherbrooke.ca\u002Fchallenge\u002Fdataset\u002F)  \n  [tip18]\n- [Tiny ImageNet](https:\u002F\u002Ftiny-imagenet.herokuapp.com\u002F)\n \n\u003Ca id=\"animals_\">\u003C\u002Fa>\n### 动物（Animals）\n\n- [Wildlife Image and Localization Dataset (species and bounding box labels)](https:\u002F\u002Flev.cs.rpi.edu\u002Fpublic\u002Fdatasets\u002Fwild.tar.gz)  \n  [wacv18]\n- [Stanford Dogs Dataset](http:\u002F\u002Fvision.stanford.edu\u002Faditya86\u002FImageNetDogs\u002F)  \n  [cvpr11]\n- [Oxford-IIIT Pet Dataset](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~vgg\u002Fdata\u002Fpets\u002F)  \n  [cvpr12]\n- [Caltech-UCSD Birds 200](http:\u002F\u002Fwww.vision.caltech.edu\u002Fvisipedia\u002FCUB-200.html) [rough segmentation] [attributes]\n- [Gold Standard Snapshot Serengeti Bounding Box Coordinates](https:\u002F\u002Fdataverse.scholarsportal.info\u002Fdataset.xhtml?persistentId=doi:10.5683\u002FSP\u002FTPB5ID)\n\n\u003Ca id=\"boundary_detectio_n_\">\u003C\u002Fa>\n## 边界检测（Boundary Detection）\n\n- [Semantic Boundaries Dataset and Benchmark](http:\u002F\u002Fhome.bharathh.info\u002Fpubs\u002Fcodes\u002FSBD\u002Fdownload.html)\n\n\u003Ca id=\"static_segmentation_\">\u003C\u002Fa>\n## 静态分割（Static Segmentation）\n\n- [COCO - Common Objects in Context](http:\u002F\u002Fcocodataset.org\u002F#download)\n- [Open Images](https:\u002F\u002Fstorage.googleapis.com\u002Fopenimages\u002Fweb\u002Findex.html)\n- [ADE20K](https:\u002F\u002Fgroups.csail.mit.edu\u002Fvision\u002Fdatasets\u002FADE20K\u002F)  \n  [cvpr17]\n- [SYNTHIA](http:\u002F\u002Fsynthia-dataset.net\u002Fdownload-2\u002F)  \n  [cvpr16]\n- [UC Berkeley Computer Vision Group - Contour Detection and Image Segmentation](https:\u002F\u002Fwww2.eecs.berkeley.edu\u002FResearch\u002FProjects\u002FCS\u002Fvision\u002Fgrouping\u002Fresources.html)\n\n\u003Ca id=\"video_segmentation_\">\u003C\u002Fa>\n## 视频分割（Video Segmentation）\n\n- [DAVIS: Densely Annotated VIdeo Segmentation](https:\u002F\u002Fdavischallenge.org\u002F)\n- [Mapillary Vistas Dataset](https:\u002F\u002Fwww.mapillary.com\u002Fdataset\u002Fvistas?pKey=0_xJqX3-c-KyTb90oG_8HQ&lat=20&lng=0&z=1.5) [street scenes] [semi-free]\n- [BDD100K](https:\u002F\u002Fbair.berkeley.edu\u002Fblog\u002F2018\u002F05\u002F30\u002Fbdd\u002F) [street scenes] [autonomous driving]\n- [ApolloScape](http:\u002F\u002Fapolloscape.auto\u002F) [street scenes] [autonomous driving]\n- [Cityscapes](https:\u002F\u002Fwww.cityscapes-dataset.com\u002F) [street scenes] [instance-level]\n- [YouTube-VOS](https:\u002F\u002Fyoutube-vos.org\u002Fdataset\u002Fvis\u002F)  \n  [iccv19]\n\n\u003Ca id=\"classificatio_n_\">\u003C\u002Fa>\n## 分类（Classification）\n\n- [ImageNet Large Scale Visual Recognition Competition 2012](http:\u002F\u002Fwww.image-net.org\u002Fchallenges\u002FLSVRC\u002F2012\u002F)\n- [Animals with Attributes 2](https:\u002F\u002Fcvml.ist.ac.at\u002FAwA2\u002F)\n- [CompCars Dataset](http:\u002F\u002Fmmlab.ie.cuhk.edu.hk\u002Fdatasets\u002Fcomp_cars\u002Findex.html)\n- [ObjectNet](https:\u002F\u002Fobjectnet.dev\u002F) [only test set]\n\n\u003Ca id=\"optical_flow_\">\u003C\u002Fa>\n## 光流（Optical Flow）\n\n- [Middlebury](http:\u002F\u002Fvision.middlebury.edu\u002Fflow\u002Fdata\u002F)\n- [MPI Sintel](http:\u002F\u002Fsintel.is.tue.mpg.de\u002F)\n- [KITTI Flow](http:\u002F\u002Fwww.cvlibs.net\u002Fdatasets\u002Fkitti\u002Feval_scene_flow.php?benchmark=flow)\n\n\u003Ca id=\"motion_prediction_\">\u003C\u002Fa>\n## 运动预测（Motion Prediction）\n\n- [Trajnet++ (A Trajectory Forecasting Challenge)](https:\u002F\u002Fwww.aicrowd.com\u002Fchallenges\u002Ftrajnet-a-trajectory-forecasting-challenge)\n- [Trajectory Forecasting Challenge](http:\u002F\u002Ftrajnet.stanford.edu\u002F)\n\n\n\u003Ca id=\"cod_e_\">\u003C\u002Fa>\n# 代码（Code）\n\n\u003Ca id=\"general_vision_\">\u003C\u002Fa>\n## 通用视觉（General Vision）\n- [Gluon CV Toolkit](https:\u002F\u002Fgithub.com\u002Fdmlc\u002Fgluon-cv) [mxnet] [pytorch]\n- [OpenMMLab Computer Vision Foundation](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmcv) [pytorch]\n\n\u003Ca id=\"multi_object_tracking__2\">\u003C\u002Fa>\n## 多目标跟踪（Multi Object Tracking）\n\n\u003Ca id=\"framework_s_\">\u003C\u002Fa>\n### 框架（Frameworks）\n* [OpenMMLab Video Perception Toolbox. It supports Video Object Detection (VID), Multiple Object Tracking (MOT), Single Object Tracking (SOT), Video Instance Segmentation (VIS) with a unified framework](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmtracking) [pytorch]\n\n\u003Ca id=\"general_\">\u003C\u002Fa>\n\n### 通用\n* [Globally-optimal greedy algorithms for tracking a variable number of objects](http:\u002F\u002Fwww.csee.umbc.edu\u002F~hpirsiav\u002Fpapers\u002Ftracking_release_v1.0.tar.gz) [cvpr11] [matlab] [[作者]](https:\u002F\u002Fwww.csee.umbc.edu\u002F~hpirsiav\u002F)    \n* [Continuous Energy Minimization for Multitarget Tracking](https:\u002F\u002Fbitbucket.org\u002Familan\u002Fcontracking) [cvpr11 \u002F iccv11 \u002F tpami 2014] [matlab]\n* [Discrete-Continuous Energy Minimization for Multi-Target Tracking](http:\u002F\u002Fwww.milanton.de\u002Ffiles\u002Fsoftware\u002Fdctracking-v1.0.zip) [cvpr12] [matlab] [[项目]](http:\u002F\u002Fwww.milanton.de\u002Fdctracking\u002Findex.html)\n* [The way they move: Tracking multiple targets with similar appearance](https:\u002F\u002Fbitbucket.org\u002Fcdicle\u002Fsmot\u002Fsrc\u002Fmaster\u002F) [iccv13] [matlab]   \n* [3D Traffic Scene Understanding from Movable Platforms](http:\u002F\u002Fwww.cvlibs.net\u002Fprojects\u002Fintersection\u002F) [[2d_tracking]](http:\u002F\u002Fwww.cvlibs.net\u002Fsoftware\u002Ftrackbydet\u002F) [pami14\u002Fkit13\u002Ficcv13\u002Fnips11] [c++\u002Fmatlab]\n* [基于无向分层关系超图的多目标跟踪](http:\u002F\u002Fwww.cbsr.ia.ac.cn\u002Fusers\u002Flywen\u002Fcodes\u002FMultiCarTracker.zip) [cvpr14] [C++] [[作者]](http:\u002F\u002Fwww.cbsr.ia.ac.cn\u002Fusers\u002Flywen\u002F)\n* [Robust online multi-object tracking based on tracklet confidence and online discriminative appearance learning](https:\u002F\u002Fdrive.google.com\u002Fopen?id=1YMqvkrVI6LOXRwcaUlAZTu_b2_5GmTAM) [cvpr14] [matlab] [(项目)](https:\u002F\u002Fsites.google.com\u002Fview\u002Finuvision\u002Fresearch)\n* [Learning to Track: Online Multi-Object Tracking by Decision Making](https:\u002F\u002Fgithub.com\u002Fyuxng\u002FMDP_Tracking) [iccv15] [matlab]\n* [Joint Tracking and Segmentation of Multiple Targets](https:\u002F\u002Fbitbucket.org\u002Familan\u002Fsegtracking) [cvpr15] [matlab]\n* [Multiple Hypothesis Tracking Revisited](http:\u002F\u002Frehg.org\u002Fmht\u002F) [iccv15] [在 MOT2015 上开源跟踪器中性能最高] [matlab]\n* [Combined Image- and World-Space Tracking in Traffic Scenes](https:\u002F\u002Fgithub.com\u002Faljosaosep\u002Fciwt) [icra 2017] [c++]\n* [Online Multi-Target Tracking with Recurrent Neural Networks](https:\u002F\u002Fbitbucket.org\u002Familan\u002Frnntracking\u002Fsrc\u002Fdefault\u002F) [aaai17] [lua\u002Ftorch7]\n* [Real-Time Multiple Object Tracking - A Study on the Importance of Speed](https:\u002F\u002Fgithub.com\u002Fsamuelmurray\u002Ftracking-by-detection) [ax1710\u002F硕士论文] [c++]        \n* [Beyond Pixels: Leveraging Geometry and Shape Cues for Online Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002FJunaidCS032\u002FMOTBeyondPixels) [icra18] [matlab]    \n* [Online Multi-Object Tracking with Dual Matching Attention Network](https:\u002F\u002Fgithub.com\u002Fjizhu1023\u002FDMAN_MOT) [eccv18] [matlab\u002Ftensorflow]    \n* [TrackR-CNN - 多目标跟踪与分割](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FTrackR-CNN) [cvpr19] [tensorflow] [[项目]](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fmots) \n* [Eliminating Exposure Bias and Metric Mismatch in Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002Fmaksay\u002Fseq-train) [cvpr19] [tensorflow]    \n* [Robust Multi-Modality Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002FZwwWayne\u002FmmMOT) [iccv19] [pytorch]    \n* [Towards Real-Time Multi-Object Tracking \u002F Joint Detection and Embedding](https:\u002F\u002Fgithub.com\u002FZhongdao\u002FTowards-Realtime-MOT) [ax1909] [pytorch] [[CMU]](https:\u002F\u002Fgithub.com\u002FJunweiLiang\u002FObject_Detection_Tracking)\n* [Deep Affinity Network for Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002FshijieS\u002FSST) [tpami19] [pytorch]    \n* [Tracking without bells and whistles](https:\u002F\u002Fgithub.com\u002Fphil-bergmann\u002Ftracking_wo_bnw) [iccv19] [pytorch]    \n* [Lifted Disjoint Paths with Application in Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002FAndreaHor\u002FLifT_Solver) [icml20] [matlab] [mot15 第1名, mot16 第3名, mot17 第2名]   \n* [Learning a Neural Solver for Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002Fdvl-tum\u002Fmot_neural_solver) [cvpr20] [pytorch] [mot15 第2名]   \n* [Tracking Objects as Points](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterTrack) [ax2004] [pytorch]\n* [Quasi-Dense Similarity Learning for Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002FSysCV\u002Fqdtrack) [ax2006] [pytorch]\n* [DEFT: Detection Embeddings for Tracking](https:\u002F\u002Fgithub.com\u002FMedChaabane\u002FDEFT) [ax2102] [pytorch]\n* [How To Train Your Deep Multi-Object Tracker](https:\u002F\u002Fgithub.com\u002FyihongXU\u002FdeepMOT) [ax1906\u002Fcvpr20] [pytorch] [[traktor\u002Fgitlab]](https:\u002F\u002Fgitlab.inria.fr\u002Fyixu\u002Fdeepmot)\n* [Track To Detect and Segment: An Online Multi-Object Tracker ](https:\u002F\u002Fgithub.com\u002FJialianW\u002FTraDeS) [cvpr21] [pytorch] [[项目]](https:\u002F\u002Fjialianwu.com\u002Fprojects\u002FTraDeS.html)\n* [MOTR: End-to-End Multiple-Object Tracking with Transformer](https:\u002F\u002Fgithub.com\u002Fmegvii-model\u002FMOTR) [ax2202] [pytorch]\n\n\n\u003Ca id=\"baselin_e__1\">\u003C\u002Fa>\n### 基线方法（Baseline）\n* [Simple Online and Realtime Tracking](https:\u002F\u002Fgithub.com\u002Fabewley\u002Fsort) [icip 2016] [python]\n* [Deep SORT : Simple Online Realtime Tracking with a Deep Association Metric](https:\u002F\u002Fgithub.com\u002Fnwojke\u002Fdeep_sort) [icip17] [python]\n* [High-Speed Tracking-by-Detection Without Using Image Information](https:\u002F\u002Fgithub.com\u002Fbochinski\u002Fiou-tracker) [avss17] [python]  \n* [A simple baseline for one-shot multi-object tracking](https:\u002F\u002Fgithub.com\u002Fifzhang\u002FFairMOT) [ax2004] [pytorch] [MOT15、16、17、20 的冠军方法]\n\n\u003Ca id=\"siamese__1\">\u003C\u002Fa>\n### Siamese 网络\n* [SiamMOT: Siamese Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002Famazon-research\u002Fsiam-mot) [ax2105] [pytorch]\n    \n\u003Ca id=\"unsupervise_d_\">\u003C\u002Fa>\n### 无监督方法（Unsupervised）\n* [Tracking by Animation: Unsupervised Learning of Multi-Object Attentive Trackers](https:\u002F\u002Fgithub.com\u002Fzhen-he\u002Ftracking-by-animation) [cvpr19] [python\u002Fc++\u002Fpytorch]\n    \n\u003Ca id=\"re_id_\">\u003C\u002Fa>\n### 重识别（Re-ID）\n* [Torchreid: Deep learning person re-identification in PyTorch](https:\u002F\u002Fgithub.com\u002FKaiyangZhou\u002Fdeep-person-reid) [ax1910] [pytorch]\n* [SMOT: Single-Shot Multi Object Tracking](https:\u002F\u002Fgithub.com\u002Fdmlc\u002Fgluon-cv\u002Ftree\u002Fmaster\u002Fgluoncv\u002Fmodel_zoo\u002Fsmot) [ax2010] [pytorch] [gluon-cv]\n* [FairMOT: On the Fairness of Detection and Re-Identification in Multiple Object Tracking](https:\u002F\u002Fgithub.com\u002Fifzhang\u002FFairMOT) [ax2004] [pytorch] [[微软]](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FFairMOT) [[BDD100K]](https:\u002F\u002Fgithub.com\u002Fdingwoai\u002FFairMOT-BDD100K) [[人脸跟踪]](https:\u002F\u002Fgithub.com\u002Fzengwb-lx\u002FFace-Tracking-usingFairMOT)\n* [Rethinking the competition between detection and ReID in Multi-Object Tracking](https:\u002F\u002Fgithub.com\u002FJudasDie\u002FSOTS) [ax2010] [pytorch] \n\n\u003Ca id=\"framework_s__1\">\u003C\u002Fa>\n#### 框架（Frameworks）\n* [PyTorch open-source toolbox for unsupervised or domain adaptive object re-ID](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002FOpenUnReID) [pytorch] \n\n\n\u003Ca id=\"graph_nn_\">\u003C\u002Fa>\n### 图神经网络（Graph NN）\n* [Joint Object Detection and Multi-Object Tracking with Graph Neural Networks](https:\u002F\u002Fgithub.com\u002Fyongxinw\u002FGSDT) [ax2006\u002F icra21] [pytorch]\n\n\u003Ca id=\"microscopy___cell_tracking__1\">\u003C\u002Fa>\n### 显微镜 \u002F 细胞跟踪（Microscopy \u002F cell tracking）\n* [Baxter Algorithms \u002F Viterbi Tracking](https:\u002F\u002Fgithub.com\u002Fklasma\u002FBaxterAlgorithms) [tmi14] [matlab]\n* [Deepcell: Accurate cell tracking and lineage construction in live-cell imaging experiments with deep learning](https:\u002F\u002Fgithub.com\u002Fvanvalenlab\u002Fdeepcell-tracking) [biorxiv1910] [tensorflow]\n\n\u003Ca id=\"3_d_\">\u003C\u002Fa>\n\n### 3D\n* [3D 多目标跟踪：一个基线方法与新的评估指标](https:\u002F\u002Fgithub.com\u002Fxinshuoweng\u002FAB3DMOT) [iros20\u002Feccvw20] [pytorch]\n* [GNN3DMOT: 基于图神经网络（Graph Neural Network, GNN）的 3D 多目标跟踪方法，结合多特征学习](https:\u002F\u002Fgithub.com\u002Fxinshuoweng\u002FGNN3DMOT) [iros20\u002Feccvw20] [pytorch]\n\n\u003Ca id=\"metrics__1\">\u003C\u002Fa>\n### 评估指标（Metrics）\n* [HOTA: 用于评估多目标跟踪的高阶指标](https:\u002F\u002Fgithub.com\u002FJonathonLuiten\u002FHOTA-metrics) [cvpr20] [python]\n\n\u003Ca id=\"single_object_tracking__2\">\u003C\u002Fa>\n## 单目标跟踪（Single Object Tracking）\n* [常见跟踪算法合集（2003–2012）](https:\u002F\u002Fgithub.com\u002Fzenhacker\u002FTrackingAlgoCollection) [c++\u002Fmatlab]\n* [商汤科技（SenseTime）单目标跟踪研究平台，实现了 SiamRPN、SiamMask 等算法](https:\u002F\u002Fgithub.com\u002FSTVIR\u002Fpysot\u002F) [pytorch]\n* [为基于颜色的无模型（model-free）跟踪正名](https:\u002F\u002Fgithub.com\u002Fjbhuang0604\u002FCF2) [cvpr15] [c++]\n* [用于视觉跟踪的层次化卷积特征](https:\u002F\u002Fgithub.com\u002Fjbhuang0604\u002FCF2) [iccv15] [matlab]\n* [基于全卷积网络（Fully Convolutional Networks）的视觉跟踪](https:\u002F\u002Fgithub.com\u002Fscott89\u002FFCNT) [iccv15] [matlab]\n* [用于视觉跟踪的层次化卷积特征](https:\u002F\u002Fgithub.com\u002Fjbhuang0604\u002FCF2) [iccv15] [matlab] \n* [DeepTracking: 利用循环神经网络（Recurrent Neural Networks）实现“超越视觉”的跟踪](https:\u002F\u002Fgithub.com\u002Fpondruska\u002FDeepTracking) [aaai16] [torch 7]\n* Learning Multi-Domain Convolutional Neural Networks for Visual Tracking [cvpr16] [vot2015 冠军] [[matlab\u002Fmatconvnet]](https:\u002F\u002Fgithub.com\u002FHyeonseobNam\u002FMDNet) [[pytorch]](https:\u002F\u002Fgithub.com\u002FHyeonseobNam\u002Fpy-MDNet)\n* [超越相关滤波器：学习用于视觉跟踪的连续卷积算子（Continuous Convolution Operators）](https:\u002F\u002Fgithub.com\u002Fmartin-danelljan\u002FContinuous-ConvOp) [eccv 2016] [matlab]\n* [用于目标跟踪的全卷积孪生网络（Fully-Convolutional Siamese Networks）](https:\u002F\u002Fgithub.com\u002Fbertinetto\u002Fsiamese-fc) [eccvw 2016] [matlab\u002Fmatconvnet] [[项目主页]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~luca\u002Fsiamese-fc.html) [[pytorch]](https:\u002F\u002Fgithub.com\u002Fhuanglianghua\u002Fsiamfc-pytorch) [[pytorch（仅训练）]](https:\u002F\u002Fgithub.com\u002Frafellerc\u002FPytorch-SiamFC)\n* [DCFNet: 用于视觉跟踪的判别相关滤波器网络（Discriminant Correlation Filters Network）](https:\u002F\u002Farxiv.org\u002Fabs\u002F1704.04057) [ax1704] [[matlab\u002Fmatconvnet]](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FDCFNet\u002F) [[pytorch]](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FDCFNet_pytorch\u002F)\n* [面向相关滤波器跟踪的端到端表示学习](https:\u002F\u002Farxiv.org\u002Fabs\u002F1704.06036) [cvpr17] [[matlab\u002Fmatconvnet]](https:\u002F\u002Fgithub.com\u002Fbertinetto\u002Fcfnet) [[tensorflow\u002F仅推理]](https:\u002F\u002Fgithub.com\u002Ftorrvision\u002Fsiamfc-tf) [[项目主页]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~luca\u002Fsiamese-fc.html)\n* [用于视觉跟踪的双深度网络（Dual Deep Network）](https:\u002F\u002Fgithub.com\u002Fchizhizhen\u002FDNT) [tip1704] [caffe]\n* [SiameseX: 用于跟踪的孪生网络（Siamese networks）简化版 PyTorch 实现，包括 SiamFC、SiamRPN、SiamRPN++、SiamVGG、SiamDW、SiamRPN-VGG](https:\u002F\u002Fgithub.com\u002Fzllrunning\u002FSiameseX.PyTorch) [pytorch]\n* [RATM: 循环注意力跟踪模型（Recurrent Attentive Tracking Model）](https:\u002F\u002Fgithub.com\u002Fsaebrahimi\u002FRATM) [cvprw17] [python]\n* [ROLO: 用于视觉目标跟踪的空间监督循环卷积神经网络（Spatially Supervised Recurrent Convolutional Neural Networks）](https:\u002F\u002Fgithub.com\u002FGuanghan\u002FROLO) [iscas 2017] [tensorflow]\n* [ECO: 高效卷积算子用于跟踪（Efficient Convolution Operators for Tracking）](https:\u002F\u002Farxiv.org\u002Fabs\u002F1611.09224) [cvpr17] [[matlab]](https:\u002F\u002Fgithub.com\u002Fmartin-danelljan\u002FECO) [[python\u002Fcuda]](https:\u002F\u002Fgithub.com\u002FStrangerZhang\u002FpyECO) [[pytorch]](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking)\n* [基于深度强化学习的动作决策网络（Action-Decision Networks）用于视觉跟踪](https:\u002F\u002Fgithub.com\u002Fildoonet\u002Ftf-adnet-tracking) [cvpr17] [tensorflow]\n* [从检测到跟踪，从跟踪到检测（Detect to Track and Track to Detect）](https:\u002F\u002Fgithub.com\u002Ffeichtenhofer\u002FDetect-Track) [iccv17] [matlab]\n* [Meta-Tracker: 视觉目标跟踪器的快速鲁棒在线自适应方法](https:\u002F\u002Fgithub.com\u002Fsilverbottlep\u002Fmeta_trackers) [eccv18] [pytorch]\n* [学习时空正则化相关滤波器用于视觉跟踪](https:\u002F\u002Fgithub.com\u002Flifeng9472\u002FSTRCF) [cvpr18] [matlab]\n* High Performance Visual Tracking with Siamese Region Proposal Network [cvpr18] [[pytorch\u002F195]](https:\u002F\u002Fgithub.com\u002Fzkisthebest\u002FSiamese-RPN) [[pytorch\u002F313]](https:\u002F\u002Fgithub.com\u002Fsongdejia\u002FSiamese-RPN-pytorch)  [[pytorch\u002Fno_train\u002F104]](https:\u002F\u002Fgithub.com\u002Fhuanglianghua\u002Fsiamrpn-pytorch) [[pytorch\u002F177]](https:\u002F\u002Fgithub.com\u002FHelloRicky123\u002FSiamese-RPN) \n* [抗干扰孪生网络（Distractor-aware Siamese Networks）用于视觉目标跟踪](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FDaSiamRPN) [eccv18] [vot18 冠军] [pytorch]\n* [VITAL: 基于对抗学习（Adversarial Learning）的视觉跟踪](https:\u002F\u002Fgithub.com\u002Fybsong00\u002FVital_release) [cvpr18] [matlab] [[pytorch]](https:\u002F\u002Fgithub.com\u002Fabnerwang\u002Fpy-Vital) [[项目主页]](https:\u002F\u002Fybsong00.github.io\u002Fcvpr18_tracking\u002Findex.html)\n* [快速在线目标跟踪与分割：一种统一方法（SiamMask）](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002FSiamMask) [cvpr19] [pytorch] [[项目主页]](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~qwang\u002FSiamMask\u002F)\n* [PyTracking: 基于 PyTorch 的通用 Python 视觉目标跟踪训练与运行框架](https:\u002F\u002Fgithub.com\u002Fvisionml\u002Fpytracking) [ECO\u002FATOM\u002FDiMP\u002FPrDiMP] [cvpr17\u002Fcvpr19\u002Ficcv19\u002Fcvpr20] [pytorch] \n* [无监督深度跟踪（Unsupervised Deep Tracking）](https:\u002F\u002Fgithub.com\u002F594422814\u002FUDT) [cvpr19] [matlab\u002Fmatconvnet] [[pytorch]](https:\u002F\u002Fgithub.com\u002F594422814\u002FUDT_pytorch)\n* [更深更宽的孪生网络用于实时视觉跟踪](https:\u002F\u002Fgithub.com\u002Fresearchmm\u002FSiamDW) [cvpr19] [pytorch]\n* [GradNet: 梯度引导网络（Gradient-Guided Network）用于视觉目标跟踪](https:\u002F\u002Fgithub.com\u002FLPXTT\u002FGradNet-Tensorflow) [iccv19] [tensorflow]\n* [`Skimming-Perusal' 跟踪：一种用于实时鲁棒长期跟踪的框架](https:\u002F\u002Fgithub.com\u002Fiiau-tracker\u002FSPLT) [iccv19] [tensorflow]\n* [学习异常抑制相关滤波器用于实时无人机（UAV）跟踪](https:\u002F\u002Fgithub.com\u002Fvision4robotics\u002FARCF-tracker) [iccv19] [matlab]\n* [学习孪生跟踪器的模型更新策略](https:\u002F\u002Fgithub.com\u002Fzhanglichao\u002Fupdatenet) [iccv19] [pytorch]\n* [SPM-Tracker: 用于实时视觉目标跟踪的串并联匹配（Series-Parallel Matching）方法](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FSPM-Tracker) [cvpr19] [pytorch] [仅推理]\n* [联合组特征选择与判别滤波器学习用于鲁棒视觉目标跟踪](https:\u002F\u002Fgithub.com\u002FXU-TIANYANG\u002FGFS-DCF) [iccv19] [matlab]\n* [Siam R-CNN: 通过重检测（Re-Detection）实现视觉跟踪](https:\u002F\u002Fgithub.com\u002FVisualComputingInstitute\u002FSiamR-CNN) [cvpr20] [tensorflow]\n* [D3S - 判别式单次分割跟踪器（Discriminative Single Shot Segmentation Tracker）](https:\u002F\u002Fgithub.com\u002Falanlukezic\u002Fd3s) [cvpr20] [pytorch\u002Fpytracking]\n* [孪生视觉跟踪的判别式鲁棒在线学习（Discriminative and Robust Online Learning）](https:\u002F\u002Fgithub.com\u002Fshallowtoil\u002FDROL) [aaai20] [pytorch\u002Fpysot]\n* [孪生框自适应网络（Siamese Box Adaptive Network）用于视觉跟踪](https:\u002F\u002Fgithub.com\u002Fhqucv\u002Fsiamban) [cvpr20] [pytorch\u002Fpysot]\n* [Ocean: 面向目标感知的无锚框跟踪（Object-aware Anchor-free Tracking）](https:\u002F\u002Fgithub.com\u002FJudasDie\u002FSOTS) [ax2010] [pytorch] \n\n\u003Ca id=\"gui_application___large_scale_tracking___animal_s_\">\u003C\u002Fa>\n\n### GUI 应用 \u002F 大规模追踪 \u002F 动物\n* [BioTracker：一个用于视觉动物追踪的开源计算机视觉框架](https:\u002F\u002Fgithub.com\u002FBioroboticsLab\u002Fbiotracker_core) [opencv\u002Fc++]\n* [Tracktor：基于图像的动物运动与行为自动追踪](https:\u002F\u002Fgithub.com\u002Fvivekhsridhar\u002Ftracktor) [opencv\u002Fc++]\n* [MARGO（Massively Automated Real-time GUI for Object-tracking）：一个用于高通量行为学（ethology）研究的平台](https:\u002F\u002Fgithub.com\u002Fde-Bivort-Lab\u002Fmargo) [matlab]\n* [idtracker.ai：在大量未标记动物群体中追踪所有个体](https:\u002F\u002Fgitlab.com\u002Fpolavieja_lab\u002Fidtrackerai) [tensorflow] [[项目主页]](https:\u002F\u002Fidtracker.ai\u002F)\n\n\u003Ca id=\"video_detectio_n__2\">\u003C\u002Fa>\n## 视频检测（Video Detection）\n* [Flow-Guided Feature Aggregation for Video Object Detection](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FFlow-Guided-Feature-Aggregation) [nips16 \u002F iccv17] [mxnet]\n* [T-CNN: Tubelets with Convolution Neural Networks](https:\u002F\u002Fgithub.com\u002Fmyfavouritekk\u002FT-CNN) [cvpr16] [python]  \n* [TPN: Tubelet Proposal Network](https:\u002F\u002Fgithub.com\u002Fmyfavouritekk\u002FTPN) [cvpr17] [python]\n* [Deep Feature Flow for Video Recognition](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FDeep-Feature-Flow) [cvpr17] [mxnet]\n* [Mobile Video Object Detection with Temporally-Aware Feature Maps](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Flstm_object_detection) [cvpr18] [Google] [tensorflow]  \n\n\u003Ca id=\"action_detectio_n_\">\u003C\u002Fa>\n### 动作检测（Action Detection）\n\u003Ca id=\"framework_s__2\">\u003C\u002Fa>\n#### 框架（Frameworks）\n+ [OpenMMLab 下一代视频理解工具箱与基准](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmaction2) [pytorch]\n\n\u003Ca id=\"static_detection_and_matching_\">\u003C\u002Fa>\n## 静态检测与匹配（Static Detection and Matching）\n\u003Ca id=\"framework_s__3\">\u003C\u002Fa>\n### 框架（Frameworks）\n+ [Tensorflow object detection API](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fobject_detection) [tensorflow]\n+ [Detectron2](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetectron2) [pytorch]\n+ [Detectron](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDetectron) [pytorch]\n+ [基于 PyTorch 的 Open MMLab 目标检测工具箱](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmdetection) [pytorch]\n+ [SimpleDet](https:\u002F\u002Fgithub.com\u002Ftusen-ai\u002Fsimpledet) [mxnet]\n\n\u003Ca id=\"region_proposal__1\">\u003C\u002Fa>\n### 区域建议（Region Proposal）   \n+ [MCG : Multiscale Combinatorial Grouping - 目标建议与分割](https:\u002F\u002Fgithub.com\u002Fjponttuset\u002Fmcg) [(项目主页)](https:\u002F\u002Fwww2.eecs.berkeley.edu\u002FResearch\u002FProjects\u002FCS\u002Fvision\u002Fgrouping\u002Fmcg\u002F) [tpami16\u002Fcvpr14] [python]\n+ [COB : Convolutional Oriented Boundaries](https:\u002F\u002Fgithub.com\u002Fkmaninis\u002FCOB) [(项目主页)](http:\u002F\u002Fwww.vision.ee.ethz.ch\u002F~cvlsegmentation\u002Fcob\u002F) [tpami18\u002Feccv16] [matlab\u002Fcaffe]\n\n\u003Ca id=\"fpn_\">\u003C\u002Fa>\n### FPN（特征金字塔网络，Feature Pyramid Networks）\n* [Feature Pyramid Networks for Object Detection](https:\u002F\u002Fgithub.com\u002Funsky\u002FFPN) [caffe\u002Fpython]  \n\n\u003Ca id=\"rcn_n__1\">\u003C\u002Fa>\n### RCNN\n* [RFCN (作者官方实现)](https:\u002F\u002Fgithub.com\u002Fdaijifeng001\u002Fr-fcn) [caffe\u002Fmatlab]\n* [RFCN-tensorflow](https:\u002F\u002Fgithub.com\u002Fxdever\u002FRFCN-tensorflow) [tensorflow]\n* [PVANet: 用于实时目标检测的轻量级深度神经网络](https:\u002F\u002Fgithub.com\u002Fsanghoon\u002Fpva-faster-rcnn) [intel] [emdnn16(nips16)]\n* Mask R-CNN [[tensorflow]](https:\u002F\u002Fgithub.com\u002FCharlesShang\u002FFastMaskRCNN) [[keras]](https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)\n* [Light-head R-CNN](https:\u002F\u002Fgithub.com\u002Fzengarden\u002Flight_head_rcnn) [cvpr18] [tensorflow]    \n* [Evolving Boxes for Fast Vehicle Detection](https:\u002F\u002Fgithub.com\u002FWilly0919\u002FEvolving_Boxes) [icme18] [caffe\u002Fpython]\n* [Cascade R-CNN (cvpr18)](http:\u002F\u002Fwww.svcl.ucsd.edu\u002Fpublications\u002Fconference\u002F2018\u002Fcvpr\u002Fcascade-rcnn.pdf) [[detectron]](https:\u002F\u002Fgithub.com\u002Fzhaoweicai\u002FDetectron-Cascade-RCNN) [[caffe]](https:\u002F\u002Fgithub.com\u002Fzhaoweicai\u002Fcascade-rcnn)  \n* [A MultiPath Network for Object Detection](https:\u002F\u002Farxiv.org\u002Fabs\u002F1604.02135) [[torch]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fmultipathnet) [bmvc16] [facebook]\n* [SNIPER: Efficient Multi-Scale Training \u002F 关于目标检测中尺度不变性的分析 - SNIP](https:\u002F\u002Fgithub.com\u002Fmahyarnajibi\u002FSNIPER) [nips18\u002Fcvpr18] [mxnet]\n\n\u003Ca id=\"ssd__1\">\u003C\u002Fa>\n### SSD（Single Shot MultiBox Detector）\n* [SSD-Tensorflow](https:\u002F\u002Fgithub.com\u002Fljanyst\u002Fssd-tensorflow) [tensorflow]\n* [SSD-Tensorflow (tf.estimator)](https:\u002F\u002Fgithub.com\u002FHiKapok\u002FSSD.TensorFlow) [tensorflow]\n* [SSD-Tensorflow (tf.slim)](https:\u002F\u002Fgithub.com\u002Fbalancap\u002FSSD-Tensorflow) [tensorflow]\n* [SSD-Keras](https:\u002F\u002Fgithub.com\u002Frykov8\u002Fssd_keras) [keras]\n* [SSD-Pytorch](https:\u002F\u002Fgithub.com\u002Famdegroot\u002Fssd.pytorch) [pytorch]\n* [Enhanced SSD with Feature Fusion and Visual Reasoning](https:\u002F\u002Fgithub.com\u002FCVlengjiaxu\u002FEnhanced-SSD-with-Feature-Fusion-and-Visual-Reasoning) [nca18] [tensorflow]\n* [RefineDet - Single-Shot Refinement Neural Network for Object Detection](https:\u002F\u002Fgithub.com\u002Fsfzhang15\u002FRefineDet) [cvpr18] [caffe]\n\n\u003Ca id=\"retinanet__1\">\u003C\u002Fa>\n### RetinaNet\n* [9.277.41](https:\u002F\u002Fgithub.com\u002Fc0nn3r\u002FRetinaNet) [pytorch]\n* [31.857.212](https:\u002F\u002Fgithub.com\u002Fkuangliu\u002Fpytorch-retinanet) [pytorch]\n* [25.274.84](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fretinanet-examples) [pytorch] [nvidia]\n* [22.869.302](https:\u002F\u002Fgithub.com\u002Fyhenon\u002Fpytorch-retinanet) [pytorch]\n\n\u003Ca id=\"yol_o__1\">\u003C\u002Fa>\n### YOLO（You Only Look Once）  \n+ [Darknet: 卷积神经网络](https:\u002F\u002Fgithub.com\u002Fpjreddie\u002Fdarknet) [c\u002Fpython]\n+ [YOLO9000: Better, Faster, Stronger - 实时目标检测，支持 9000 个类别！](https:\u002F\u002Fgithub.com\u002Fphilipperemy\u002Fyolo-9000) [c\u002Fpython]\n+ [Darkflow](https:\u002F\u002Fgithub.com\u002Fthtrieu\u002Fdarkflow) [tensorflow]\n+ [Pytorch Yolov2 ](https:\u002F\u002Fgithub.com\u002Fmarvis\u002Fpytorch-yolo2) [pytorch]\n+ [适用于 Windows 和 Linux 的 Yolo-v3 和 Yolo-v2](https:\u002F\u002Fgithub.com\u002FAlexeyAB\u002Fdarknet) [c\u002Fpython]\n+ [YOLOv3 in PyTorch](https:\u002F\u002Fgithub.com\u002Fultralytics\u002Fyolov3) [pytorch]\n+ [pytorch-yolo-v3 ](https:\u002F\u002Fgithub.com\u002Fayooshkathuria\u002Fpytorch-yolo-v3) [pytorch] [无训练功能] [[教程]](https:\u002F\u002Fblog.paperspace.com\u002Fhow-to-implement-a-yolo-object-detector-in-pytorch\u002F)\n+ [YOLOv3_TensorFlow](https:\u002F\u002Fgithub.com\u002Fwizyoung\u002FYOLOv3_TensorFlow) [tensorflow]\n+ [tensorflow-yolo-v3](https:\u002F\u002Fgithub.com\u002Fmystic123\u002Ftensorflow-yolo-v3) [tensorflow slim]\n+ [tensorflow-yolov3](https:\u002F\u002Fgithub.com\u002FYunYang1994\u002Ftensorflow-yolov3) [tensorflow slim]\n+ [keras-yolov3](https:\u002F\u002Fgithub.com\u002Fqqwweee\u002Fkeras-yolo3) [keras]  \n+ YOLOv4 [[darknet - c\u002Fpython]](https:\u002F\u002Fgithub.com\u002FAlexeyAB\u002Fdarknet) [[tensorflow]](https:\u002F\u002Fgithub.com\u002Fhunglc007\u002Ftensorflow-yolov4-tflite) [[pytorch\u002F711]](https:\u002F\u002Fgithub.com\u002FWongKinYiu\u002FPyTorch_YOLOv4) [[pytorch\u002FONNX\u002FTensorRT\u002F1.9k]](https:\u002F\u002Fgithub.com\u002FTianxiaomo\u002Fpytorch-YOLOv4) [[pytorch 3D]](https:\u002F\u002Fgithub.com\u002Fmaudzung\u002FComplex-YOLOv4-Pytorch)\n+ [YOLOv5](https:\u002F\u002Fgithub.com\u002Fultralytics\u002Fyolov5) [pytorch] \n+ [YOLOX](https:\u002F\u002Fgithub.com\u002FMegvii-BaseDetection\u002FYOLOX) [pytorch] [MegEngine](https:\u002F\u002Fgithub.com\u002FMegEngine\u002FYOLOX) [ax2107]\n\n\u003Ca id=\"anchor_free__1\">\u003C\u002Fa>\n\n### 无锚框（Anchor Free）\n* [FoveaBox: 超越基于锚框的目标检测器](https:\u002F\u002Fgithub.com\u002Ftaokong\u002FFoveaBox) [ax1904] [pytorch\u002Fmmdetection]\n* [Cornernet: 将目标检测视为成对关键点检测](https:\u002F\u002Fgithub.com\u002Fprinceton-vl\u002FCornerNet) [ax1903\u002Feccv18] [pytorch]\n* [FCOS: 全卷积单阶段目标检测](https:\u002F\u002Fgithub.com\u002Ftianzhi0549\u002FFCOS) [iccv19] [pytorch] [[VoVNet]](https:\u002F\u002Fgithub.com\u002Fvov-net\u002FVoVNet-FCOS) [[HRNet]](https:\u002F\u002Fgithub.com\u002FHRNet\u002FHRNet-FCOS) [[NAS]](https:\u002F\u002Fgithub.com\u002FLausannen\u002FNAS-FCOS) [[FCOS_PLUS]](https:\u002F\u002Fgithub.com\u002Fyqyao\u002FFCOS_PLUS)\n* [用于单阶段目标检测的特征选择无锚框模块](https:\u002F\u002Fgithub.com\u002Fhdjang\u002FFeature-Selective-Anchor-Free-Module-for-Single-Shot-Object-Detection) [cvpr19] [pytorch]\n* [CenterNet: 将目标视为点](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FCenterNet) [ax1904] [pytorch]\n* [通过聚合极值点与中心点实现自底向上的目标检测](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FExtremeNet) [cvpr19] [pytorch]\n* [RepPoints：用于目标检测的点集表示](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FRepPoints) [iccv19] [pytorch] [microsoft]\n* [DETR: 基于 Transformer 的端到端目标检测](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetr) [ax200528] [pytorch] [facebook]\n* [通过自适应训练样本选择弥合基于锚框与无锚框检测之间的差距](https:\u002F\u002Fgithub.com\u002Fsfzhang15\u002FATSS) [cvpr20] [pytorch]\n\n\u003Ca id=\"mis_c__2\">\u003C\u002Fa>\n### 其他（Misc）\n* [用于目标检测的关系网络（Relation Networks）](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FRelation-Networks-for-Object-Detection) [cvpr18] [mxnet]\n* [DeNet: 使用定向稀疏采样的可扩展实时目标检测](https:\u002F\u002Fgithub.com\u002Flachlants\u002Fdenet) [iccv17(poster)] [theano]\n* [用于目标检测的多尺度位置感知核表示](https:\u002F\u002Fgithub.com\u002FHwang64\u002FMLKP) [cvpr18] [caffe\u002Fpython]\n\n\u003Ca id=\"matchin_g_\">\u003C\u002Fa>\n### 匹配（Matching）  \n+ [Matchnet](https:\u002F\u002Fgithub.com\u002Fhanxf\u002Fmatchnet)\n+ [通过训练卷积神经网络比较图像块实现立体匹配](https:\u002F\u002Fgithub.com\u002Fjzbontar\u002Fmc-cnn)\n\n\u003Ca id=\"boundary_detectio_n__1\">\u003C\u002Fa>\n### 边界检测（Boundary Detection）  \n+ [整体嵌套边缘检测（Holistically-Nested Edge Detection, HED）(iccv15)](https:\u002F\u002Fgithub.com\u002Fs9xie\u002Fhed) [caffe]       \n+ [使用深度学习进行边缘检测（HED）](https:\u002F\u002Fgithub.com\u002FAkuanchang\u002FEdge-Detection-using-Deep-Learning) [tensorflow]\n+ [OpenCV 中的整体嵌套边缘检测（HED）](https:\u002F\u002Fgithub.com\u002Fopencv\u002Fopencv\u002Fblob\u002Fmaster\u002Fsamples\u002Fdnn\u002Fedge_detection.py) [python\u002Fc++]       \n+ [使用逐点互信息实现清晰边界检测（eccv14）](https:\u002F\u002Fgithub.com\u002Fphillipi\u002Fcrisp-boundaries) [matlab]\n+ [密集极值 Inception 网络：面向鲁棒边缘检测的 CNN 模型](https:\u002F\u002Fgithub.com\u002Fphillipi\u002Fcrisp-boundaries) [wacv20] [tensorflow](https:\u002F\u002Fgithub.com\u002Fxavysp\u002FDexiNed\u002Ftree\u002Fmaster\u002Flegacy) [pytorch](https:\u002F\u002Fgithub.com\u002Fxavysp\u002FDexiNed)\n\n\u003Ca id=\"text_detectio_n_\">\u003C\u002Fa>\n### 文本检测（Text Detection）  \n+ [使用可微分二值化的实时场景文本检测](https:\u002F\u002Fgithub.com\u002FMhLiao\u002FDB) [pytorch] [aaai20] \n\n\u003Ca id=\"framework_s__4\">\u003C\u002Fa>\n#### 框架（Frameworks）\n+ [OpenMMLab 文本检测、识别与理解工具箱](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmocr) [pytorch]\n\n\u003Ca id=\"3d_detectio_n_\">\u003C\u002Fa>\n### 3D 检测（3D Detection）  \n\u003Ca id=\"framework_s__5\">\u003C\u002Fa>\n#### 框架（Frameworks）\n+ [OpenMMLab 下一代通用 3D 目标检测平台](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmdetection3d) [pytorch]\n+ [基于 LiDAR 的 3D 目标检测工具箱 OpenPCDet](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002FOpenPCDet) [pytorch]\n\n\u003Ca id=\"optical_flow__1\">\u003C\u002Fa>\n## 光流（Optical Flow）\n* [FlowNet 2.0: 使用深度网络进行光流估计的演进 (cvpr17)](https:\u002F\u002Farxiv.org\u002Fabs\u002F1612.01925) [[caffe]](https:\u002F\u002Fgithub.com\u002Flmb-freiburg\u002Fflownet2) [[pytorch\u002Fnvidia]](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fflownet2-pytorch)\n* [SPyNet: 用于光流的空间金字塔网络 (cvpr17)](https:\u002F\u002Farxiv.org\u002Fabs\u002F1702.02295) [[lua]](https:\u002F\u002Fgithub.com\u002Fanuragranj\u002Fspynet) [[pytorch]](https:\u002F\u002Fgithub.com\u002Fsniklaus\u002Fpytorch-spynet)\n* [引导式光流学习 (cvprw17)](https:\u002F\u002Farxiv.org\u002Fabs\u002F1702.02295) [[caffe]](https:\u002F\u002Fgithub.com\u002Fbryanyzhu\u002FGuidedNet) [[tensorflow]](https:\u002F\u002Fgithub.com\u002Fbryanyzhu\u002FdeepOF)\n* [使用密集逆搜索（DIS）的快速光流](https:\u002F\u002Fgithub.com\u002Ftikroeger\u002FOF_DIS) [eccv16] [C++]\n* [一种用于计算实时光流的滤波器方法](https:\u002F\u002Fgithub.com\u002Fjadarve\u002Foptical-flow-filter) [ral16] [c++\u002Fcuda - matlab, python 封装]\n* [PatchBatch - 一种用于光流的批增强损失](https:\u002F\u002Fgithub.com\u002FDediGadot\u002FPatchBatch) [cvpr16] [python\u002Ftheano]\n* [分段刚性场景流（Piecewise Rigid Scene Flow）](https:\u002F\u002Fgithub.com\u002Fvogechri\u002FPRSM) [iccv13\u002Feccv14\u002Fijcv15] [c++\u002Fmatlab]\n* [DeepFlow v2](https:\u002F\u002Farxiv.org\u002Fabs\u002F1611.00850) [iccv13] [[c++\u002Fpython\u002Fmatlab]](https:\u002F\u002Fgithub.com\u002Fzimenglan-sysu-512\u002Fdeep-flow), [[项目主页]](http:\u002F\u002Flear.inrialpes.fr\u002Fsrc\u002Fdeepflow\u002F)\n* [光流数据代价的评估](https:\u002F\u002Fgithub.com\u002Fvogechri\u002FDataFlow) [gcpr13] [matlab]\n\n\u003Ca id=\"framework_s__6\">\u003C\u002Fa>\n### 框架（Frameworks）\n+ [OpenMMLab 光流工具箱与基准](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmflow) [pytorch]\n\n\n\u003Ca id=\"instance_segmentation_\">\u003C\u002Fa>\n## 实例分割（Instance Segmentation）\n* [全卷积实例感知语义分割](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FFCIS) [cvpr17] [coco16 冠军] [mxnet]\n* [通过多任务网络级联实现实例感知语义分割](https:\u002F\u002Fgithub.com\u002Fdaijifeng001\u002FMNC) [cvpr16] [caffe] [coco15 冠军]    \n* [DeepMask\u002FSharpMask](https:\u002F\u002Farxiv.org\u002Fabs\u002F1603.08695) [nips15\u002Feccv16] [facebook] [[torch]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdeepmask) [[tensorflow]](https:\u002F\u002Fgithub.com\u002Faby2s\u002Fsharpmask)  [[pytorch\u002Fdeepmask]](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002Fdeepmask-pytorch\u002F) \n* [同时检测与分割（Simultaneous Detection and Segmentation）](https:\u002F\u002Fgithub.com\u002Fbharath272\u002Fsds_eccv2014) [eccv14] [matlab] [[项目主页]](https:\u002F\u002Fwww2.eecs.berkeley.edu\u002FResearch\u002FProjects\u002FCS\u002Fvision\u002Fshape\u002Fsds\u002F)    \n* [PANet](https:\u002F\u002Fgithub.com\u002FShuLiu1993\u002FPANet) [cvpr18] [pytorch]\n* [RetinaMask](https:\u002F\u002Fgithub.com\u002Fchengyangfu\u002Fretinamask) [arxviv1901] [pytorch]\n* [Mask Scoring R-CNN](https:\u002F\u002Fgithub.com\u002Fzjhuang22\u002Fmaskscoring_rcnn) [cvpr19] [pytorch]\n* [DeepMAC](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Fblob\u002Fmaster\u002Fresearch\u002Fobject_detection\u002Fg3doc\u002Fdeepmac.md) [ax2104] [tensorflow]\n* [Swin Transformer](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FSwin-Transformer) [iccv21] [pytorch] [microsoft]\n\n\u003Ca id=\"framework_s__7\">\u003C\u002Fa>\n### 框架（Frameworks）\n* [PyTorch 中实例分割与目标检测算法的快速、模块化参考实现](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fmaskrcnn-benchmark) [pytorch] [facebook]\n* [PaddleDetection，基于 PaddlePaddle 的目标检测与实例分割工具箱](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleDetection) [2019]\n\n\u003Ca id=\"semantic_segmentation_\">\u003C\u002Fa>\n\n## 语义分割（Semantic Segmentation）\n* [Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation](https:\u002F\u002Fgithub.com\u002Fswamiviv\u002FLSD-seg) [cvpr18] [spotlight] [pytorch]\n* [Few-shot Segmentation Propagation with Guided Networks](https:\u002F\u002Fgithub.com\u002Fshelhamer\u002Frevolver) [ax1806] [pytorch] [incomplete]\n* [Pytorch-segmentation-toolbox](https:\u002F\u002Fgithub.com\u002Fspeedinghzl\u002Fpytorch-segmentation-toolbox) [DeeplabV3 和 PSPNet] [pytorch]\n* [DeepLab](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fdeeplab) [tensorflow]\n* [Auto-DeepLab](https:\u002F\u002Fgithub.com\u002FMenghaoGuo\u002FAutoDeeplab) [pytorch]\n* [DeepLab v3+](https:\u002F\u002Fgithub.com\u002Fjfzhang95\u002Fpytorch-deeplab-xception) [pytorch]\n* [Deep Extreme Cut (DEXTR): From Extreme Points to Object Segmentation](https:\u002F\u002Fgithub.com\u002Fscaelles\u002FDEXTR-PyTorch)[cvpr18][[project]](https:\u002F\u002Fcvlsegmentation.github.io\u002Fdextr\u002F) [pytorch]\n* [FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation](https:\u002F\u002Fgithub.com\u002Fwuhuikai\u002FFastFCN)[ax1903][[project]](http:\u002F\u002Fwuhuikai.me\u002FFastFCNProject\u002F) [pytorch]\n\n\u003Ca id=\"framework_s__8\">\u003C\u002Fa>\n### 框架（Frameworks）\n+ [OpenMMLab 语义分割工具箱与基准（OpenMMLab Semantic Segmentation Toolbox and Benchmark）](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmsegmentation) [pytorch]\n\n\u003Ca id=\"polyp_\">\u003C\u002Fa>\n### 息肉分割（Polyp）\n* [PraNet: Parallel Reverse Attention Network for Polyp Segmentation](https:\u002F\u002Fgithub.com\u002FDengPingFan\u002FPraNet)[miccai20]\n* [PHarDNet-MSEG: A Simple Encoder-Decoder Polyp Segmentation Neural Network that Achieves over 0.9 Mean Dice and 86 FPS](https:\u002F\u002Fgithub.com\u002Fjames128333\u002FHarDNet-MSEG)[ax2101]\n\n\u003Ca id=\"panoptic_segmentation_\">\u003C\u002Fa>\n## 全景分割（Panoptic Segmentation）\n* [Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetectron2\u002Ftree\u002Fmain\u002Fprojects\u002FPanoptic-DeepLab) [cvpr20] [pytorch]\n\n\u003Ca id=\"video_segmentation__1\">\u003C\u002Fa>\n## 视频分割（Video Segmentation）\n* [Improving Semantic Segmentation via Video Prediction and Label Relaxation](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fsemantic-segmentation) [cvpr19] [pytorch] [nvidia]\n* [PReMVOS: Proposal-generation, Refinement and Merging for Video Object Segmentation](https:\u002F\u002Fgithub.com\u002FJonathonLuiten\u002FPReMVOS) [accv18\u002Fcvprw18\u002Feccvw18] [tensorflow]\n* [MaskTrackRCNN for video instance segmentation](https:\u002F\u002Fgithub.com\u002Fyoutubevos\u002FMaskTrackRCNN) [iccv19] [pytorch\u002Fdetectron]\n* [MaskTrackRCNN](https:\u002F\u002Fgithub.com\u002Fyoutubevos\u002FMaskTrackRCNN) [iccv19] [pytorch\u002Fdetectron]\n* [Video Instance Segmentation using Inter-Frame Communication Transformers](https:\u002F\u002Fgithub.com\u002Fsukjunhwang\u002FIFC) [nips21] [pytorch\u002Fdetectron]\n* [VNext: SeqFormer \u002F IDOL](https:\u002F\u002Fgithub.com\u002Fwjf5203\u002FVNext) [eccv22] [pytorch\u002Fdetectron2]\n* [SeqFormer: Sequential Transformer for Video Instance Segmentation](https:\u002F\u002Fgithub.com\u002Fwjf5203\u002FSeqFormer) [eccv22] [pytorch\u002Fdetectron2]\n* [VITA: Video Instance Segmentation via Object Token Association](https:\u002F\u002Fgithub.com\u002Fsukjunhwang\u002Fvita) [nips22] [pytorch\u002Fdetectron2]\n\n\u003Ca id=\"panoptic_video_segmentation_\">\u003C\u002Fa>\n### 全景视频分割（Panoptic Video Segmentation）\n* [ViP-DeepLab](https:\u002F\u002Fgithub.com\u002Fjoe-siyuan-qiao\u002FViP-DeepLab) [cvpr21] \n\n\u003Ca id=\"motion_prediction__1\">\u003C\u002Fa>\n## 运动预测（Motion Prediction）\n* [Self-Supervised Learning via Conditional Motion Propagation](https:\u002F\u002Fgithub.com\u002FXiaohangZhan\u002Fconditional-motion-propagation) [cvpr19] [pytorch]\n* [A Neural Temporal Model for Human Motion Prediction](https:\u002F\u002Fgithub.com\u002Fcr7anand\u002Fneural_temporal_models) [cvpr19] [tensorflow]   \n* [Learning Trajectory Dependencies for Human Motion Prediction](https:\u002F\u002Fgithub.com\u002Fwei-mao-2019\u002FLearnTrajDep) [iccv19] [pytorch]   \n* [Structural-RNN: Deep Learning on Spatio-Temporal Graphs](https:\u002F\u002Fgithub.com\u002Fzhaolongkzz\u002Fhuman_motion) [cvpr15] [tensorflow]   \n* [A Keras multi-input multi-output LSTM-based RNN for object trajectory forecasting](https:\u002F\u002Fgithub.com\u002FMarlonCajamarca\u002FKeras-LSTM-Trajectory-Prediction) [keras]   \n* [Transformer Networks for Trajectory Forecasting](https:\u002F\u002Fgithub.com\u002FFGiuliari\u002FTrajectory-Transformer) [ax2003] [pytorch]  \n* [Regularizing neural networks for future trajectory prediction via IRL framework](https:\u002F\u002Fgithub.com\u002Fd1024choi\u002Ftraj-pred-irl) [ietcv1907] [tensorflow]  \n* [Peeking into the Future: Predicting Future Person Activities and Locations in Videos](https:\u002F\u002Fgithub.com\u002FJunweiLiang\u002Fnext-prediction) [cvpr19] [tensorflow]  \n* [DAG-Net: Double Attentive Graph Neural Network for Trajectory Forecasting](https:\u002F\u002Fgithub.com\u002Falexmonti19\u002Fdagnet) [ax200526] [pytorch]  \n* [MCENET: Multi-Context Encoder Network for Homogeneous Agent Trajectory Prediction in Mixed Traffic](https:\u002F\u002Fgithub.com\u002Fsugmichaelyang\u002FMCENET) [ax200405] [tensorflow]  \n* [Human Trajectory Prediction in Socially Interacting Crowds Using a CNN-based Architecture](https:\u002F\u002Fgithub.com\u002Fbiy001\u002Fsocial-cnn-pytorch) [pytorch]  \n* [A tool set for trajectory prediction, ready for pip install](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FTrajectory-Prediction-Tools) [icai19\u002Fwacv19]  [pytorch]  \n* [RobustTP: End-to-End Trajectory Prediction for Heterogeneous Road-Agents in Dense Traffic with Noisy Sensor Inputs](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FTrajectory-Prediction-Tools) [acmcscs19]  [pytorch\u002Ftensorflow]  \n* [The Garden of Forking Paths: Towards Multi-Future Trajectory Prediction](https:\u002F\u002Fgithub.com\u002FJunweiLiang\u002FMultiverse) [cvpr20] [dummy] \n* [Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Prediction](https:\u002F\u002Fgithub.com\u002Flmb-freiburg\u002FMultimodal-Future-Prediction) [cvpr19] [tensorflow] \n* [Adversarial Loss for Human Trajectory Prediction](https:\u002F\u002Fgithub.com\u002Fvita-epfl\u002FAdversarialLoss-SGAN) [hEART19] [pytorch] \n* [Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks](https:\u002F\u002Fgithub.com\u002Fagrimgupta92\u002Fsgan) [cvpr18] [pytorch] \n* [Forecasting Trajectory and Behavior of Road-Agents Using Spectral Clustering in Graph-LSTMs](https:\u002F\u002Fgithub.com\u002Frohanchandra30\u002FSpectral-Trajectory-and-Behavior-Prediction) [ax1912] [pytorch] \n* [Study of attention mechanisms for trajectory prediction in Deep Learning](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FTrajectory-Prediction-Tools) [msc thesis]  [python]  \n* [A python implementation of multi-model estimation algorithm for trajectory tracking and prediction, research project from BMW ABSOLUT self-driving bus project.](https:\u002F\u002Fgithub.com\u002FchrisHuxi\u002FTrajectory_Predictor) [python]  \n* [Predicting Human Trajectories](https:\u002F\u002Fgithub.com\u002Fkarthik4444\u002Fnn-trajectory-prediction) [theano]  \n* [Implementation of Recurrent Neural Networks for future trajectory prediction of pedestrians](https:\u002F\u002Fgithub.com\u002Faroongta\u002FPedestrian_Trajectory_Prediction) [pytorch]  \n\n\u003Ca id=\"pose_estimation_\">\u003C\u002Fa>\n## 姿态估计（Pose Estimation）\n\u003Ca id=\"framework_s__9\">\u003C\u002Fa>\n### 框架（Frameworks）\n+ [OpenMMLab 姿态估计工具箱与基准（OpenMMLab Pose Estimation Toolbox and Benchmark）](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmpose) [pytorch]\n\n\u003Ca id=\"autoencoder_s_\">\u003C\u002Fa>\n\n## 自编码器（Autoencoders）\n* [β-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework](https:\u002F\u002Fopenreview.net\u002Fforum?id=Sy2fzU9gl) [iclr17] [deepmind] [[tensorflow]](https:\u002F\u002Fgithub.com\u002Fmiyosuda\u002Fdisentangled_vae) [[tensorflow]](https:\u002F\u002Fgithub.com\u002FLynnHo\u002FVAE-Tensorflow) [[pytorch]](https:\u002F\u002Fgithub.com\u002F1Konny\u002FBeta-VAE)\n* [Disentangling by Factorising](https:\u002F\u002Fgithub.com\u002F1Konny\u002FFactorVAE) [ax1806] [pytorch]   \n\n\u003Ca id=\"classificatio_n__1\">\u003C\u002Fa>\n## 分类（Classification）\n* [Learning Efficient Convolutional Networks Through Network Slimming](https:\u002F\u002Fgithub.com\u002Fmiyosuda\u002Fasync_deep_reinforce) [iccv17] [pytorch]\n\n\u003Ca id=\"framework_s__10\">\u003C\u002Fa>\n### 框架（Frameworks）\n+ [OpenMMLab 图像分类工具箱与基准测试](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmclassification) [pytorch]\n\n\u003Ca id=\"deep_rl_\">\u003C\u002Fa>\n## 深度强化学习（Deep RL）\n* [Asynchronous Methods for Deep Reinforcement Learning ](https:\u002F\u002Fgithub.com\u002Fmiyosuda\u002Fasync_deep_reinforce)\n\n\u003Ca id=\"annotatio_n_\">\u003C\u002Fa>\n## 标注（Annotation）\n- [LabelImg](https:\u002F\u002Fgithub.com\u002Ftzutalin\u002FlabelImg)\n- [ByLabel: A Boundary Based Semi-Automatic Image Annotation Tool](https:\u002F\u002Fgithub.com\u002FNathanUA\u002FByLabel)\n- [Bounding Box Editor and Exporter](https:\u002F\u002Fgithub.com\u002Fpersts\u002FBBoxEE)\n- [VGG Image Annotator](http:\u002F\u002Fwww.robots.ox.ac.uk\u002F~vgg\u002Fsoftware\u002Fvia\u002F)\n- [Visual Object Tagging Tool: 一个用于从图像和视频构建端到端目标检测模型的 Electron 应用](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002FVoTT)\n- [PixelAnnotationTool](https:\u002F\u002Fgithub.com\u002Fabreheret\u002FPixelAnnotationTool)\n- [labelme：使用 Python 进行图像多边形标注（支持多边形、矩形、圆形、线条、点及图像级标签标注）](https:\u002F\u002Fgithub.com\u002Fwkentaro\u002Flabelme)\n- [VATIC - 来自加州尔湾的视频标注工具](https:\u002F\u002Fgithub.com\u002Fcvondrick\u002Fvatic) [ijcv12] [[项目主页]](http:\u002F\u002Fwww.cs.columbia.edu\u002F~vondrick\u002Fvatic\u002F)\n- [计算机视觉标注工具（CVAT）](https:\u002F\u002Fgithub.com\u002Fopencv\u002Fcvat)\n- [图像标注工具](https:\u002F\u002Fbitbucket.org\u002Fueacomputervision\u002Fimage-labelling-tool\u002F)\n- [Labelbox](https:\u002F\u002Fgithub.com\u002FLabelbox\u002FLabelbox) [付费]\n- [RectLabel：一款用于边界框目标检测和分割的图像标注工具](https:\u002F\u002Frectlabel.com\u002F) [付费]\n- [Onepanel：面向生产级视觉 AI 的平台，集成了模型构建、自动标注、数据处理和模型训练流水线等完整组件](https:\u002F\u002Fgithub.com\u002Fonepanelio\u002Fcore) [[文档]](https:\u002F\u002Fdocs.onepanel.ai\u002Fdocs\u002Fgetting-started\u002Fquickstart\u002F)\n\n\u003Ca id=\"editing_\">\u003C\u002Fa>\n### 编辑（Editing）\n- [OpenMMLab 图像与视频编辑工具箱](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmediting)\n\n\u003Ca id=\"augmentatio_n_\">\u003C\u002Fa>\n### 数据增强（Augmentation）\n- [Augmentor：用于机器学习的 Python 图像增强库](https:\u002F\u002Fgithub.com\u002Fmdbloice\u002FAugmentor)\n- [Albumentations：快速图像增强库，并提供对其他库的易用封装](https:\u002F\u002Fgithub.com\u002Falbumentations-team\u002Falbumentations)\n- [imgaug：用于机器学习实验的图像增强库](https:\u002F\u002Fgithub.com\u002Faleju\u002Fimgaug)\n- [solt：基于轻量级数据变换的图像流处理库](https:\u002F\u002Fgithub.com\u002FMIPT-Oulu\u002Fsolt)\n\n\u003Ca id=\"deep_learning__2\">\u003C\u002Fa>\n## 深度学习（Deep Learning）\n* [可变形卷积网络（Deformable Convolutional Networks）](https:\u002F\u002Fgithub.com\u002Fmsracver\u002FDeformable-ConvNets)\n* [RNNexp](https:\u002F\u002Fgithub.com\u002Fasheshjain399\u002FRNNexp)\n* [Grad-CAM：梯度加权类激活映射（Gradient-weighted Class Activation Mapping）](https:\u002F\u002Fgithub.com\u002Framprs\u002Fgrad-cam\u002F)\n\n\u003Ca id=\"class_imbalanc_e_\">\u003C\u002Fa>\n### 类别不平衡（Class Imbalance）\n* [Imbalanced Dataset Sampler](https:\u002F\u002Fgithub.com\u002Fufoym\u002Fimbalanced-dataset-sampler) [pytorch]\n* [PyTorch 中的可迭代数据集重采样](https:\u002F\u002Fgithub.com\u002FMaxHalford\u002Fpytorch-resample) [pytorch]\n\n\u003Ca id=\"few_shot_learning_\">\u003C\u002Fa>\n### 小样本学习（Few shot learning）\n* [OpenMMLab 小样本学习工具箱与基准测试](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002Fmmfewshot) [pytorch]\n\n\u003Ca id=\"unsupervised_learning__2\">\u003C\u002Fa>\n### 无监督学习（Unsupervised learning）\n* [自监督学习工具箱与基准测试（Self-Supervised Learning Toolbox and Benchmark）](https:\u002F\u002Fgithub.com\u002Fopen-mmlab\u002FOpenSelfSup) [pytorch]\n\n\u003Ca id=\"collections_\">\u003C\u002Fa>\n# 资源集合（Collections）\n\n\u003Ca id=\"dataset_s__1\">\u003C\u002Fa>\n## 数据集（Datasets）\n* [Awesome Public Datasets](https:\u002F\u002Fgithub.com\u002Fawesomedata\u002Fawesome-public-datasets) \n* [交通监控数据集列表](https:\u002F\u002Fgithub.com\u002Fgustavovelascoh\u002Ftraffic-surveillance-dataset) \n* [机器学习数据集：来自网络的大型机器学习数据集列表](https:\u002F\u002Fwww.datasetlist.com\u002F) \n* [亚历山大标注信息图书馆：生物学与保护领域](http:\u002F\u002Flila.science\u002Fdatasets) [[其他保护相关数据集]](http:\u002F\u002Flila.science\u002Fotherdatasets) \n* [THOTH：数据集与图像](https:\u002F\u002Fthoth.inrialpes.fr\u002Fdata) \n* [Google AI 数据集](https:\u002F\u002Fai.google\u002Ftools\u002Fdatasets\u002F) \n* [Google Cloud Storage 公共数据集](https:\u002F\u002Fcloud.google.com\u002Fstorage\u002Fdocs\u002Fpublic-datasets\u002F) \n* [微软研究院开放数据](https:\u002F\u002Fmsropendata.com\u002F) \n* [Earth Engine 数据目录](https:\u002F\u002Fdevelopers.google.com\u002Fearth-engine\u002Fdatasets\u002Fcatalog\u002F) \n* [AWS 开放数据注册表](https:\u002F\u002Fregistry.opendata.aws\u002F) \n* [Kaggle 数据集](https:\u002F\u002Fwww.kaggle.com\u002Fdatasets) \n* [CVonline：图像数据库](http:\u002F\u002Fhomepages.inf.ed.ac.uk\u002Frbf\u002FCVonline\u002FImagedbase.htm) \n* [计算机视觉合成数据：计算机视觉领域的合成数据集与工具列表](https:\u002F\u002Fgithub.com\u002Funrealcv\u002Fsynthetic-computer-vision) \n* [pgram 机器学习数据集](https:\u002F\u002Fpgram.com\u002Fcategory\u002Fvision\u002F) \n* [pgram 视觉数据集](https:\u002F\u002Fpgram.com\u002F) \n\n\u003Ca id=\"deep_learning__3\">\u003C\u002Fa>\n## 深度学习（Deep Learning）\n- [Model Zoo：发现开源深度学习代码与预训练模型](https:\u002F\u002Fmodelzoo.co\u002F)\n\n\u003Ca id=\"static_detectio_n__2\">\u003C\u002Fa>\n## 静态检测（Static Detection）\n- [基于深度学习的目标检测](https:\u002F\u002Fhandong1587.github.io\u002Fdeep_learning\u002F2015\u002F10\u002F09\u002Fobject-detection.html)\n\n\u003Ca id=\"video_detectio_n__3\">\u003C\u002Fa>\n## 视频检测（Video Detection）\n- [基于深度学习的视频目标检测](https:\u002F\u002Fhandong1587.github.io\u002Fdeep_learning\u002F2015\u002F10\u002F09\u002Fobject-detection.html#video-object-detection)\n\n\u003Ca id=\"single_object_tracking__3\">\u003C\u002Fa>\n## 单目标跟踪（Single Object Tracking）\n- [视觉跟踪论文列表](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002Fbenchmark_results)\n- [基于深度学习的跟踪论文列表](https:\u002F\u002Fgithub.com\u002Fhandong1587\u002Fhandong1587.github.io\u002Fblob\u002Fmaster\u002F_posts\u002Fdeep_learning\u002F2015-10-09-tracking.md)\n- [在 OTB 上评估的单目标跟踪器列表](https:\u002F\u002Fgithub.com\u002Ffoolwood\u002Fbenchmark_results)\n- [基于相关滤波（Correlation Filter）的跟踪器集合，包含论文、代码等链接](https:\u002F\u002Fgithub.com\u002Flukaswals\u002Fcf-trackers)\n- [VOT2018 跟踪器仓库](http:\u002F\u002Fwww.votchallenge.net\u002Fvot2018\u002Ftrackers.html)\n- [CUHK 数据集](http:\u002F\u002Fmmlab.ie.cuhk.edu.hk.login.ezproxy.library.ualberta.ca\u002Fdatasets.html)\n- [CVPR19 视觉跟踪论文总结](https:\u002F\u002Flinkinpark213.com\u002F2019\u002F06\u002F11\u002Fcvpr19-track\u002F )\n- [单目标视觉跟踪器](https:\u002F\u002Fgithub.com\u002Fczla\u002Fdaily-paper-visual-tracking)\n    \n\u003Ca id=\"multi_object_tracking__3\">\u003C\u002Fa>\n\n## 多目标跟踪（Multi Object Tracking, MOT）\n* [多目标跟踪论文列表](http:\u002F\u002Fperception.yale.edu\u002FBrian\u002FrefGuides\u002FMOT.html)   \n* [近年来多目标跟踪（MOT）论文合集，附有笔记](https:\u002F\u002Fgithub.com\u002Fhuanglianghua\u002Fmot-papers)  \n* [Papers with Code：多目标跟踪](https:\u002F\u002Fpaperswithcode.com\u002Ftask\u002Fmultiple-object-tracking\u002Fcodeless)  \n* [多目标跟踪论文列表及源代码](https:\u002F\u002Fgithub.com\u002FSpyderXu\u002Fmulti-object-tracking-paper-list)  \n\n\u003Ca id=\"static_segmentation__1\">\u003C\u002Fa>\n## 静态分割（Static Segmentation）\n* [分割相关论文与代码](https:\u002F\u002Fhandong1587.github.io\u002Fdeep_learning\u002F2015\u002F10\u002F09\u002Fsegmentation.html)  \n* [Segmentation.X：语义分割（semantic segmentation）、实例分割（instance segmentation）、全景分割（panoptic segmentation）和视频分割（video segmentation）的论文与基准](https:\u002F\u002Fgithub.com\u002FwutianyiRosun\u002FSegmentation.X) \n* [带代码的实例分割论文](https:\u002F\u002Fpaperswithcode.com\u002Ftask\u002Finstance-segmentation) \n\n\u003Ca id=\"video_segmentation__2\">\u003C\u002Fa>\n## 视频分割（Video Segmentation）\n* [YouTube-VIS 验证集上的视频实例分割](https:\u002F\u002Fpaperswithcode.com\u002Fsota\u002Fvideo-instance-segmentation-on-youtube-vis-1?p=seqformer-a-frustratingly-simple-model-for)\n\n\n\u003Ca id=\"motion_prediction__2\">\u003C\u002Fa>\n## 运动预测（Motion Prediction）\n* [Awesome-Trajectory-Prediction](https:\u002F\u002Fgithub.com\u002Fxuehaouwa\u002FAwesome-Trajectory-Prediction\u002Fblob\u002Fmaster\u002FREADME.md)  \n* [Awesome Interaction-aware Behavior and Trajectory Prediction（交互感知的行为与轨迹预测）](https:\u002F\u002Fgithub.com\u002Fjiachenli94\u002FAwesome-Interaction-aware-Trajectory-Prediction)  \n* [人类轨迹预测数据集](https:\u002F\u002Fgithub.com\u002Famiryanj\u002FOpenTraj)  \n\n\u003Ca id=\"deep_compressed_sensin_g_\">\u003C\u002Fa>\n## 深度压缩感知（Deep Compressed Sensing）\n* [可复现的深度压缩感知](https:\u002F\u002Fgithub.com\u002FAtenaKid\u002FReproducible-Deep-Compressive-Sensing)  \n\n\u003Ca id=\"mis_c__3\">\u003C\u002Fa>\n## 其他（Misc）\n* [Papers With Code：机器学习最新进展](https:\u002F\u002Fpaperswithcode.com\u002F)\n* [Awesome Deep Ecology（深度生态学资源）](https:\u002F\u002Fgithub.com\u002Fpatrickcgray\u002Fawesome-deep-ecology)\n* [Matlab 框架、库和软件列表](https:\u002F\u002Fgithub.com\u002Fuhub\u002Fawesome-matlab)\n* [人脸识别（Face Recognition）](https:\u002F\u002Fgithub.com\u002FChanChiChoi\u002Fawesome-Face_Recognition)\n* [一个月的机器学习论文摘要](https:\u002F\u002Fmedium.com\u002F@hyponymous\u002Fa-month-of-machine-learning-paper-summaries-ddd4dcf6cfa5)\n* [Awesome-model-compression-and-acceleration（模型压缩与加速）](https:\u002F\u002Fgithub.com\u002Fmemoiry\u002FAwesome-model-compression-and-acceleration\u002Fblob\u002Fmaster\u002FREADME.md)\n* [Model-Compression-Papers（模型压缩论文）](https:\u002F\u002Fgithub.com\u002Fchester256\u002FModel-Compression-Papers)\n\n\u003Ca id=\"tutorials_\">\u003C\u002Fa>\n# 教程（Tutorials）\n\n\u003Ca id=\"collections__1\">\u003C\u002Fa>\n## 资源合集（Collections）\n* [PyTorch 深度学习教程](https:\u002F\u002Fgithub.com\u002Fsgrvinod\u002FDeep-Tutorials-for-PyTorch)\n\n\u003Ca id=\"multi_object_tracking__4\">\u003C\u002Fa>\n## 多目标跟踪（Multi Object Tracking）\n* [什么是多目标跟踪（MOT）系统？](https:\u002F\u002Fdeepomatic.com\u002Fen\u002Fmoving-beyond-deepomatic-learns-how-to-track-multiple-objects\u002F)\n\n\u003Ca id=\"static_detectio_n__3\">\u003C\u002Fa>\n## 静态检测（Static Detection）\n* [使用 Transformer 实现端到端目标检测](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fend-to-end-object-detection-with-transformers)\n* [深度学习目标检测全面综述](https:\u002F\u002Ftowardsdatascience.com\u002Fdeep-learning-for-object-detection-a-comprehensive-review-73930816d8d9)\n* [目标检测深度学习算法综述](https:\u002F\u002Fmedium.com\u002Fcomet-app\u002Freview-of-deep-learning-algorithms-for-object-detection-c1f3d437b852)  \n* [Inception 网络各版本简易指南](https:\u002F\u002Ftowardsdatascience.com\u002Fa-simple-guide-to-the-versions-of-the-inception-network-7fc52b863202)    \n* [R-CNN、Fast R-CNN、Faster R-CNN、YOLO — 目标检测算法](https:\u002F\u002Ftowardsdatascience.com\u002Fr-cnn-fast-r-cnn-faster-r-cnn-yolo-object-detection-algorithms-36d53571365e)\n* [深度学习目标检测入门指南](https:\u002F\u002Fwww.pyimagesearch.com\u002F2018\u002F05\u002F14\u002Fa-gentle-guide-to-deep-learning-object-detection\u002F)\n* [RetinaNet 背后的直觉](https:\u002F\u002Fmedium.com\u002F@14prakash\u002Fthe-intuition-behind-retinanet-eb636755607d)\n* [YOLO——“你只看一次”，实时目标检测详解](https:\u002F\u002Ftowardsdatascience.com\u002Fyolo-you-only-look-once-real-time-object-detection-explained-492dc9230006)\n* [理解用于目标检测的特征金字塔网络（FPN）](https:\u002F\u002Fmedium.com\u002F@jonathan_hui\u002Funderstanding-feature-pyramid-networks-for-object-detection-fpn-45b227b9106c)\n* [在 Keras 上使用 SqueezeDet 实现快速目标检测](https:\u002F\u002Fmedium.com\u002Fomnius\u002Ffast-object-detection-with-squeezedet-on-keras-5cdd124b46ce)\n* [感兴趣区域池化（RoI Pooling）详解](https:\u002F\u002Fdeepsense.ai\u002Fregion-of-interest-pooling-explained\u002F)\n\n\u003Ca id=\"video_detectio_n__4\">\u003C\u002Fa>\n## 视频检测（Video Detection）\n* [微软如何做视频目标检测——在一个模型中统一视频目标检测架构中的最佳技术](https:\u002F\u002Fmedium.com\u002Fnurture-ai\u002Fhow-microsoft-does-video-object-detection-unifying-the-best-techniques-in-video-object-detection-b78b63e3f1d8)\n\n\u003Ca id=\"instance_segmentation__1\">\u003C\u002Fa>\n## 实例分割（Instance Segmentation）\n* [色彩点缀：使用 Mask R-CNN 和 TensorFlow 实现实例分割](https:\u002F\u002Fengineering.matterport.com\u002Fsplash-of-color-instance-segmentation-with-mask-r-cnn-and-tensorflow-7c761e238b46)\n* [Mask R-CNN 简明理解](https:\u002F\u002Fmedium.com\u002F@alittlepain833\u002Fsimple-understanding-of-mask-rcnn-134b5b330e95)\n* [学会分割](https:\u002F\u002Fresearch.fb.com\u002Fblog\u002F2016\u002F08\u002Flearning-to-segment\u002F)\n* [分析 Facebook 计算机视觉方法背后的论文](https:\u002F\u002Fadeshpande3.github.io\u002FAnalyzing-the-Papers-Behind-Facebook's-Computer-Vision-Approach\u002F)\n* [论文回顾：MNC —— 2015 COCO 分割冠军的多任务网络级联](https:\u002F\u002Ftowardsdatascience.com\u002Freview-mnc-multi-task-network-cascade-winner-in-2015-coco-segmentation-instance-segmentation-42a9334e6a34)\n* [论文回顾：FCIS —— 2016 COCO 分割冠军](https:\u002F\u002Ftowardsdatascience.com\u002Freview-fcis-winner-in-2016-coco-segmentation-instance-segmentation-ee2d61f465e2)\n* [论文回顾：InstanceFCN —— 实例敏感得分图](https:\u002F\u002Ftowardsdatascience.com\u002Freview-instancefcn-instance-sensitive-score-maps-instance-segmentation-dbfe67d4ee92)\n\n\u003Ca id=\"deep_learning__4\">\u003C\u002Fa>\n## 深度学习（Deep Learning）\n\n\u003Ca id=\"optimizatio_n_\">\u003C\u002Fa>\n### 优化（Optimization）\n* [学习率调度（Learning Rate Scheduling）](https:\u002F\u002Fwww.deeplearningwizard.com\u002Fdeep_learning\u002Fboosting_models_pytorch\u002Flr_scheduling\u002F)\n\n\u003Ca id=\"class_imbalanc_e__1\">\u003C\u002Fa>\n\n### 类别不平衡（Class Imbalance）\n* [从不平衡数据中学习](https:\u002F\u002Fwww.jeremyjordan.me\u002Fimbalanced-data\u002F)\n* [从不平衡类别中学习](https:\u002F\u002Fwww.svds.com\u002Flearning-imbalanced-classes\u002F)\n* [在机器学习中处理不平衡数据集](https:\u002F\u002Ftowardsdatascience.com\u002Fhandling-imbalanced-datasets-in-machine-learning-7a0e84220f28) [medium]\n* [如何处理类别不平衡问题](https:\u002F\u002Fmedium.com\u002Fquantyca\u002Fhow-to-handle-class-imbalance-problem-9ee3062f2499) [medium]\n* [处理不平衡数据的方法](https:\u002F\u002Ftowardsdatascience.com\u002Fmethods-for-dealing-with-imbalanced-data-5b761be45a18) [towardsdatascience]\n* [如何在机器学习中处理不平衡类别](https:\u002F\u002Felitedatascience.com\u002Fimbalanced-classes) [elitedatascience]\n* [处理不平衡数据的 7 种技术](https:\u002F\u002Fwww.kdnuggets.com\u002F2017\u002F06\u002F7-techniques-handle-imbalanced-data.html) [kdnuggets]\n* [处理机器学习中不平衡类别的 10 种技术](https:\u002F\u002Fwww.analyticsvidhya.com\u002Fblog\u002F2020\u002F07\u002F10-techniques-to-deal-with-class-imbalance-in-machine-learning\u002F) [analyticsvidhya]\n\n\u003Ca id=\"rnn__2\">\u003C\u002Fa>\n## RNN（循环神经网络，Recurrent Neural Networks）\n* [循环神经网络的惊人有效性](http:\u002F\u002Fkarpathy.github.io\u002F2015\u002F05\u002F21\u002Frnn-effectiveness\u002F)\n* [理解 LSTM 网络](http:\u002F\u002Fcolah.github.io\u002Fposts\u002F2015-08-Understanding-LSTMs\u002F)\n\n\u003Ca id=\"deep_rl__1\">\u003C\u002Fa>\n## 深度强化学习（Deep RL，Deep Reinforcement Learning）\n* [深度强化学习：从像素玩 Pong 游戏](http:\u002F\u002Fkarpathy.github.io\u002F2016\u002F05\u002F31\u002Frl\u002F)\n* [揭秘深度强化学习](https:\u002F\u002Fwww.intelnervana.com\u002Fdemystifying-deep-reinforcement-learning\u002F)\n\n\u003Ca id=\"autoencoder_s__1\">\u003C\u002Fa>\n## 自编码器（Autoencoders）\n* [自编码器指南](https:\u002F\u002Fyaledatascience.github.io\u002F2016\u002F10\u002F29\u002Fautoencoders.html)\n* [应用深度学习 - 第三部分：自编码器](https:\u002F\u002Ftowardsdatascience.com\u002Fapplied-deep-learning-part-3-autoencoders-1c083af4d798)\n* [去噪自编码器（Denoising Autoencoders）](http:\u002F\u002Fdeeplearning.net\u002Ftutorial\u002FdA.html)\n* [堆叠去噪自编码器（Stacked Denoising Autoencoders）](https:\u002F\u002Fskymind.ai\u002Fwiki\u002Fstacked-denoising-autoencoder)\n* [LSTM 自编码器入门简介](https:\u002F\u002Fmachinelearningmastery.com\u002Flstm-autoencoders\u002F)\n* [使用 TensorFlow 实现变分自编码器（Variational Autoencoder）](https:\u002F\u002Fjmetzen.github.io\u002F2015-11-27\u002Fvae.html)\n* [使用 TensorFlow Probability 层实现变分自编码器](https:\u002F\u002Fmedium.com\u002Ftensorflow\u002Fvariational-autoencoders-with-tensorflow-probability-layers-d06c658931b7)\n\n\n\u003Ca id=\"blogs_\">\u003C\u002Fa>\n# 博客（Blogs）\n\n* [Facebook AI](https:\u002F\u002Fai.facebook.com\u002Fblog\u002F)\n* [Google AI](https:\u002F\u002Fai.googleblog.com\u002F)\n* [Google DeepMind](https:\u002F\u002Fdeepmind.com\u002Fblog)\n* [Deep Learning Wizard](https:\u002F\u002Fwww.deeplearningwizard.com\u002F)\n* [Towards Data Science](https:\u002F\u002Ftowardsdatascience.com\u002F)\n* [Jay Alammar：一次一个概念可视化机器学习](https:\u002F\u002Fjalammar.github.io\u002F)\n* [Inside Machine Learning：关于机器学习、云计算和数据的深度文章，由 IBM 精选](https:\u002F\u002Fmedium.com\u002Finside-machine-learning)\n* [colah's blog](http:\u002F\u002Fcolah.github.io\u002F)\n* [Jeremy Jordan](https:\u002F\u002Fwww.jeremyjordan.me\u002F)\n* [Silicon Valley Data Science](https:\u002F\u002Fwww.svds.com\u002Ftag\u002Fdata-science\u002F)\n* [Illarion’s Notes](https:\u002F\u002Fikhlestov.github.io\u002Fpages\u002F)","# Deep-Learning-for-Tracking-and-Detection 快速上手指南\n\n## 环境准备\n\n本项目为资源集合型仓库，不包含可直接运行的主程序，但提供了大量论文、数据集与代码链接。若需运行其中引用的模型（如 YOLO、Mask R-CNN 等），建议准备以下环境：\n\n- **操作系统**：Linux \u002F macOS \u002F Windows（推荐 Ubuntu 20.04+）\n- **Python 版本**：≥ 3.7\n- **深度学习框架**（按需安装）：\n  - PyTorch ≥ 1.8 或 TensorFlow ≥ 2.5\n  - CUDA ≥ 11.0（如使用 GPU）\n- **基础依赖**：\n  ```bash\n  pip install numpy opencv-python matplotlib tqdm\n  ```\n\n> 💡 国内用户建议使用清华源加速安装：\n> ```bash\n> pip install -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple 包名\n> ```\n\n## 安装步骤\n\n本项目本身无需安装，克隆仓库即可浏览资源：\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fyour-repo\u002FDeep-Learning-for-Tracking-and-Detection.git\ncd Deep-Learning-for-Tracking-and-Detection\n```\n\n如需运行具体模型（以 **YOLOv4** 为例）：\n\n```bash\n# 克隆官方实现（推荐 AlexeyAB\u002Fdarknet）\ngit clone https:\u002F\u002Fgithub.com\u002FAlexeyAB\u002Fdarknet\ncd darknet\nmake  # Linux\u002FmacOS 编译（需已安装 OpenCV 和 CUDA）\n```\n\n或使用 PyTorch 版本（如 ultralytics\u002Fyolov5）：\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fultralytics\u002Fyolov5\ncd yolov5\npip install -r requirements.txt -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n```\n\n## 基本使用\n\n### 示例：使用 YOLOv5 进行目标检测\n\n```python\nimport torch\n\n# 加载预训练模型\nmodel = torch.hub.load('ultralytics\u002Fyolov5', 'yolov5s', pretrained=True)\n\n# 推理图像\nresults = model('https:\u002F\u002Fultralytics.com\u002Fimages\u002Fzidane.jpg')\n\n# 显示结果\nresults.show()\n```\n\n### 示例：加载 Mask R-CNN（Keras 实现）\n\n```python\nfrom mrcnn import model as modellib\nfrom mrcnn.config import Config\n\nclass InferenceConfig(Config):\n    NAME = \"coco\"\n    GPU_COUNT = 1\n    IMAGES_PER_GPU = 1\n    NUM_CLASSES = 1 + 80  # COCO has 80 classes\n\nconfig = InferenceConfig()\nmodel = modellib.MaskRCNN(mode=\"inference\", config=config, model_dir=\".\u002Flogs\")\nmodel.load_weights(\"mask_rcnn_coco.h5\", by_name=True)\n```\n\n> 📌 提示：所有论文 PDF、代码链接、数据集信息均已在仓库对应目录中整理，可直接访问 `static_detection\u002F`、`multi_object_tracking\u002F` 等子目录查看。","某生物医学研究团队正在开发一套用于活细胞动态行为分析的自动化系统，需对显微镜视频中的多个细胞进行长时间、高精度的追踪与分裂事件检测。\n\n### 没有 Deep-Learning-for-Tracking-and-Detection 时\n- 团队需手动搜索最新论文和开源代码，耗费大量时间筛选适用于细胞追踪的模型，信息分散且版本混乱。\n- 缺乏针对显微成像特点（如低对比度、密集重叠）的专用数据集参考，导致模型泛化能力差。\n- 多目标追踪与检测模块需从零搭建，难以复现前沿方法（如基于图神经网络或Re-ID的关联策略）。\n- 评估指标不统一，无法客观比较不同算法在细胞分裂、遮挡等复杂场景下的性能。\n- 团队成员在YOLO、RetinaNet等静态检测框架之间反复试错，缺乏系统性技术路线指导。\n\n### 使用 Deep-Learning-for-Tracking-and-Detection 后\n- 通过项目中“Microscopy \u002F Cell Tracking”分类快速定位适配的论文、数据集（如Cell Tracking Challenge）和代码实现，大幅缩短调研周期。\n- 直接复用整理好的多目标追踪基线模型（如基于Siamese网络或Graph NN的方案），并结合显微场景微调，提升追踪连续性。\n- 利用项目提供的标准化评估指标（如MOTA、IDF1）对算法进行量化对比，精准优化分裂事件检测逻辑。\n- 借助“Static Detection”和“Multi Object Tracking”下的主流框架（如YOLOv5 + DeepSORT组合），快速构建端到端流水线。\n- 团队聚焦于生物学问题本身，而非底层技术整合，研发效率显著提升。\n\nDeep-Learning-for-Tracking-and-Detection 将碎片化的学术资源系统化，让科研团队能高效落地复杂的视觉追踪任务。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fabhineet123_Deep-Learning-for-Tracking-and-Detection_a9cc33a8.png","abhineet123","Abhineet Singh","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fabhineet123_5070e3a4.jpg",null,"University of Alberta","Edmonton","http:\u002F\u002Fwebdocs.cs.ualberta.ca\u002F~asingh1\u002F","https:\u002F\u002Fgithub.com\u002Fabhineet123",[84,88],{"name":85,"color":86,"percentage":87},"HTML","#e34c26",100,{"name":89,"color":90,"percentage":91},"Shell","#89e051",0,2502,654,"2026-03-28T07:35:54",4,"未说明",{"notes":98,"python":96,"dependencies":99},"该仓库主要为论文、数据集和相关资源的集合，并未提供统一的代码实现或安装说明。具体运行环境需参考各子项目（如 Mask R-CNN、YOLOv4、SNIPER 等）各自的官方代码库要求。部分链接指向 GitHub 项目（如 matterport\u002FMask_RCNN、AlexeyAB\u002Fdarknet），建议查阅对应项目的 README 获取详细依赖和环境配置信息。",[],[13,14],[102,103,104,105,106,107,108,109,110,111],"deep-learning","object-detection","detection","tracking-by-detection","tracking","papers","paper-collection","code-collection","segmentation","optical-flow",10,"2026-03-27T02:49:30.150509","2026-04-06T08:52:30.967578",[116],{"id":117,"question_zh":118,"answer_zh":119,"source_url":120},105,"DeepSORT 是否属于多目标跟踪（MOT）方法？","是的，DeepSORT 属于多目标跟踪（MOT）方法。项目维护者已确认并修正了相关内容。","https:\u002F\u002Fgithub.com\u002Fabhineet123\u002FDeep-Learning-for-Tracking-and-Detection\u002Fissues\u002F2",[]]