[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-balavenkatesh3322--CV-pretrained-model":3,"tool-balavenkatesh3322--CV-pretrained-model":61},[4,18,26,36,44,53],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":10,"last_commit_at":24,"category_tags":25,"status":17},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,"2026-04-05T11:01:52",[14,15,13],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":32,"last_commit_at":33,"category_tags":34,"status":17},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",148568,2,"2026-04-09T23:34:24",[14,13,35],"语言模型",{"id":37,"name":38,"github_repo":39,"description_zh":40,"stars":41,"difficulty_score":32,"last_commit_at":42,"category_tags":43,"status":17},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。",108111,"2026-04-08T11:23:26",[14,15,13],{"id":45,"name":46,"github_repo":47,"description_zh":48,"stars":49,"difficulty_score":32,"last_commit_at":50,"category_tags":51,"status":17},6121,"gemini-cli","google-gemini\u002Fgemini-cli","gemini-cli 是一款由谷歌推出的开源 AI 命令行工具，它将强大的 Gemini 大模型能力直接集成到用户的终端环境中。对于习惯在命令行工作的开发者而言，它提供了一条从输入提示词到获取模型响应的最短路径，无需切换窗口即可享受智能辅助。\n\n这款工具主要解决了开发过程中频繁上下文切换的痛点，让用户能在熟悉的终端界面内直接完成代码理解、生成、调试以及自动化运维任务。无论是查询大型代码库、根据草图生成应用，还是执行复杂的 Git 操作，gemini-cli 都能通过自然语言指令高效处理。\n\n它特别适合广大软件工程师、DevOps 人员及技术研究人员使用。其核心亮点包括支持高达 100 万 token 的超长上下文窗口，具备出色的逻辑推理能力；内置 Google 搜索、文件操作及 Shell 命令执行等实用工具；更独特的是，它支持 MCP（模型上下文协议），允许用户灵活扩展自定义集成，连接如图像生成等外部能力。此外，个人谷歌账号即可享受免费的额度支持，且项目基于 Apache 2.0 协议完全开源，是提升终端工作效率的理想助手。",100752,"2026-04-10T01:20:03",[52,13,15,14],"插件",{"id":54,"name":55,"github_repo":56,"description_zh":57,"stars":58,"difficulty_score":32,"last_commit_at":59,"category_tags":60,"status":17},4721,"markitdown","microsoft\u002Fmarkitdown","MarkItDown 是一款由微软 AutoGen 团队打造的轻量级 Python 工具，专为将各类文件高效转换为 Markdown 格式而设计。它支持 PDF、Word、Excel、PPT、图片（含 OCR）、音频（含语音转录）、HTML 乃至 YouTube 链接等多种格式的解析，能够精准提取文档中的标题、列表、表格和链接等关键结构信息。\n\n在人工智能应用日益普及的今天，大语言模型（LLM）虽擅长处理文本，却难以直接读取复杂的二进制办公文档。MarkItDown 恰好解决了这一痛点，它将非结构化或半结构化的文件转化为模型“原生理解”且 Token 效率极高的 Markdown 格式，成为连接本地文件与 AI 分析 pipeline 的理想桥梁。此外，它还提供了 MCP（模型上下文协议）服务器，可无缝集成到 Claude Desktop 等 LLM 应用中。\n\n这款工具特别适合开发者、数据科学家及 AI 研究人员使用，尤其是那些需要构建文档检索增强生成（RAG）系统、进行批量文本分析或希望让 AI 助手直接“阅读”本地文件的用户。虽然生成的内容也具备一定可读性，但其核心优势在于为机器",93400,"2026-04-06T19:52:38",[52,14],{"id":62,"github_repo":63,"name":64,"description_en":65,"description_zh":66,"ai_summary_zh":67,"readme_en":68,"readme_zh":69,"quickstart_zh":70,"use_case_zh":71,"hero_image_url":72,"owner_login":73,"owner_name":74,"owner_avatar_url":75,"owner_bio":76,"owner_company":77,"owner_location":78,"owner_email":79,"owner_twitter":80,"owner_website":81,"owner_url":82,"languages":77,"stars":83,"forks":84,"last_commit_at":85,"license":86,"difficulty_score":87,"env_os":88,"env_gpu":89,"env_ram":89,"env_deps":90,"category_tags":98,"github_topics":101,"view_count":32,"oss_zip_url":77,"oss_zip_packed_at":77,"status":17,"created_at":121,"updated_at":122,"faqs":123,"releases":124},6108,"balavenkatesh3322\u002FCV-pretrained-model","CV-pretrained-model","A collection of computer vision pre-trained models.","CV-pretrained-model 是一个专为计算机视觉领域打造的开源模型资源库，旨在帮助开发者和研究人员快速找到高质量的预训练模型。在人工智能开发中，从零开始训练一个图像识别或目标检测模型往往需要耗费大量时间、算力和数据。CV-pretrained-model 通过汇集基于 TensorFlow、Keras、PyTorch、Caffe 及 MXNet 等主流框架的成熟模型（如 YOLO、Mask R-CNN、MobileNet 等），让用户可以直接复用他人已在大规模数据集上训练好的成果，将其作为解决类似问题的起点，从而大幅降低开发门槛并提升效率。\n\n该项目不仅提供了涵盖物体定位、实例分割、语义分割及实时检测等多种任务的模型列表，还详细标注了每个模型的描述、适用框架及开源许可证信息，方便用户根据需求灵活选择。此外，项目推荐结合 Netron 工具可视化查看网络架构，帮助用户更深入地理解模型结构。无论是希望快速构建自动驾驶图像算法的工程师，还是从事学术研究的科研人员，亦或是想要尝试 AI 应用的原型设计师，都能在这里找到合适的“基石”，避免重复造轮子，将精力更多地集中在业务逻辑与创新","CV-pretrained-model 是一个专为计算机视觉领域打造的开源模型资源库，旨在帮助开发者和研究人员快速找到高质量的预训练模型。在人工智能开发中，从零开始训练一个图像识别或目标检测模型往往需要耗费大量时间、算力和数据。CV-pretrained-model 通过汇集基于 TensorFlow、Keras、PyTorch、Caffe 及 MXNet 等主流框架的成熟模型（如 YOLO、Mask R-CNN、MobileNet 等），让用户可以直接复用他人已在大规模数据集上训练好的成果，将其作为解决类似问题的起点，从而大幅降低开发门槛并提升效率。\n\n该项目不仅提供了涵盖物体定位、实例分割、语义分割及实时检测等多种任务的模型列表，还详细标注了每个模型的描述、适用框架及开源许可证信息，方便用户根据需求灵活选择。此外，项目推荐结合 Netron 工具可视化查看网络架构，帮助用户更深入地理解模型结构。无论是希望快速构建自动驾驶图像算法的工程师，还是从事学术研究的科研人员，亦或是想要尝试 AI 应用的原型设计师，都能在这里找到合适的“基石”，避免重复造轮子，将精力更多地集中在业务逻辑与创新上。","\n![Maintenance](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FMaintained%3F-YES-green.svg)\n![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FRelease-PROD-yellow.svg)\n![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLanguages-MULTI-blue.svg)\n![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-lightgrey.svg)\n\n# Computer Vision Pretrained Models\n\n![CV logo](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fbalavenkatesh3322_CV-pretrained-model_readme_dd3d9bf50810.jpg)\n\n## What is pre-trained Model?\nA pre-trained model is a model created by some one else to solve a similar problem. Instead of building a model from scratch to solve a similar problem, we can use the model trained on other problem as a starting point. A pre-trained model may not be 100% accurate in your application.\n\nFor example, if you want to build a self learning car. You can spend years to build a decent image recognition algorithm from scratch or you can take inception model (a pre-trained model) from Google which was built on [ImageNet](http:\u002F\u002Fwww.image-net.org\u002F) data to identify images in those pictures.\n\n## Other Pre-trained Models\n* [NLP Pre-trained Models](https:\u002F\u002Fgithub.com\u002Fbalavenkatesh3322\u002FNLP-pretrained-model).\n* [Audio and Speech Pre-trained Models](https:\u002F\u002Fgithub.com\u002Fbalavenkatesh3322\u002Faudio-pretrained-model)\n\n## Model Deployment library\n* [Model Serving](https:\u002F\u002Fgithub.com\u002Fbalavenkatesh3322\u002Fmodel_deployment)\n\n### Framework\n\n* [Tensorflow](#tensorflow)\n* [Keras](#keras)\n* [PyTorch](#pytorch)\n* [Caffe](#caffe)\n* [MXNet](#mxnet)\n\n### Model visualization\nYou can see visualizations of each model's network architecture by using [Netron](https:\u002F\u002Fgithub.com\u002Flutzroeder\u002FNetron).\n\n![CV logo](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fbalavenkatesh3322_CV-pretrained-model_readme_76d47d615285.png)\n\n### Tensorflow \u003Ca name=\"tensorflow\"\u002F>\n\n| Model Name | Description | Framework | License |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [ObjectDetection]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fobject_detection)  | Localizing and identifying multiple objects in a single image.| `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [Mask R-CNN]( https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)  | The model generates bounding boxes and segmentation masks for each instance of an object in the image. It's based on Feature Pyramid Network (FPN) and a ResNet101 backbone.     | `Tensorflow`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmatterport\u002FMask_RCNN\u002Fmaster\u002FLICENSE )\n| [Faster-RCNN]( https:\u002F\u002Fgithub.com\u002Fsmallcorgi\u002FFaster-RCNN_TF)  | This is an experimental Tensorflow implementation of Faster RCNN - a convnet for object detection with a region proposal network.     | `Tensorflow`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fsmallcorgi\u002FFaster-RCNN_TF\u002Fmaster\u002FLICENSE )\n| [YOLO TensorFlow]( https:\u002F\u002Fgithub.com\u002Fgliese581gg\u002FYOLO_tensorflow)  | This is tensorflow implementation of the YOLO:Real-Time Object Detection.     | `Tensorflow`| [Custom]( https:\u002F\u002Fraw.githubusercontent.com\u002Fgliese581gg\u002FYOLO_tensorflow\u002Fmaster\u002FLICENSE )\n| [YOLO TensorFlow ++]( https:\u002F\u002Fgithub.com\u002Fthtrieu\u002Fdarkflow)  | TensorFlow implementation of 'YOLO: Real-Time Object Detection', with training and an actual support for real-time running on mobile devices.     | `Tensorflow`| [GNU GENERAL PUBLIC LICENSE]( https:\u002F\u002Fraw.githubusercontent.com\u002Fthtrieu\u002Fdarkflow\u002Fmaster\u002FLICENSE )\n| [MobileNet]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Fblob\u002Fmaster\u002Fresearch\u002Fslim\u002Fnets\u002Fmobilenet_v1.md)  | MobileNets trade off between latency, size and accuracy while comparing favorably with popular models from the literature.     | `Tensorflow`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [DeepLab]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fdeeplab)  | Deep labeling for semantic image segmentation.     | `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [Colornet]( https:\u002F\u002Fgithub.com\u002Fpavelgonchar\u002Fcolornet)  | Neural Network to colorize grayscale images.     | `Tensorflow`| Not Found\n| [SRGAN]( https:\u002F\u002Fgithub.com\u002Ftensorlayer\u002Fsrgan)  | Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network.    | `Tensorflow`| Not Found\n| [DeepOSM]( https:\u002F\u002Fgithub.com\u002Ftrailbehind\u002FDeepOSM)  | Train TensorFlow neural nets with OpenStreetMap features and satellite imagery.     | `Tensorflow`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftrailbehind\u002FDeepOSM\u002Fmaster\u002FLICENSE )\n| [Domain Transfer Network]( https:\u002F\u002Fgithub.com\u002Fyunjey\u002Fdomain-transfer-network)  | Implementation of Unsupervised Cross-Domain Image Generation.  | `Tensorflow`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fyunjey\u002Fdomain-transfer-network\u002Fmaster\u002FLICENSE )\n| [Show, Attend and Tell]( https:\u002F\u002Fgithub.com\u002Fyunjey\u002Fshow-attend-and-tell)  | Attention Based Image Caption Generator.     | `Tensorflow`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fyunjey\u002Fshow-attend-and-tell\u002Fmaster\u002FLICENSE )\n| [android-yolo]( https:\u002F\u002Fgithub.com\u002Fnatanielruiz\u002Fandroid-yolo)  | Real-time object detection on Android using the YOLO network, powered by TensorFlow.    | `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fnatanielruiz\u002Fandroid-yolo\u002Fmaster\u002FLICENSE )\n| [DCSCN Super Resolution]( https:\u002F\u002Fgithub.com\u002Fjiny2001\u002Fdcscn-super-resolutiont)  | This is a tensorflow implementation of \"Fast and Accurate Image Super Resolution by Deep CNN with Skip Connection and Network in Network\", a deep learning based Single-Image Super-Resolution (SISR) model.     | `Tensorflow`| Not Found\n| [GAN-CLS]( https:\u002F\u002Fgithub.com\u002Fzsdonghao\u002Ftext-to-image)  | This is an experimental tensorflow implementation of synthesizing images.     | `Tensorflow`| Not Found\n| [U-Net]( https:\u002F\u002Fgithub.com\u002Fzsdonghao\u002Fu-net-brain-tumor)  | For Brain Tumor Segmentation.     | `Tensorflow`| Not Found\n| [Improved CycleGAN]( https:\u002F\u002Fgithub.com\u002Fluoxier\u002FCycleGAN_Tensorlayer)  |Unpaired Image to Image Translation.     | `Tensorflow`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fluoxier\u002FCycleGAN_Tensorlayer\u002Fmaster\u002FLICENSE )\n| [Im2txt]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fim2txt)  | Image-to-text neural network for image captioning.     | `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [SLIM]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fslim)  | Image classification models in TF-Slim.     | `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [DELF]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fdelf)  | Deep local features for image matching and retrieval.     | `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [Compression]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fcompression)  | Compressing and decompressing images using a pre-trained Residual GRU network.     | `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [AttentionOCR]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fattention_ocr)  | A model for real-world image text extraction.     | `Tensorflow`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ Back To Top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n### Keras \u003Ca name=\"keras\"\u002F>\n\n| Model Name | Description | Framework | License |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [Mask R-CNN]( https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)  | The model generates bounding boxes and segmentation masks for each instance of an object in the image. It's based on Feature Pyramid Network (FPN) and a ResNet101 backbone.| `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmatterport\u002FMask_RCNN\u002Fmaster\u002FLICENSE )\n| [VGG16]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fvgg16.py)  | Very Deep Convolutional Networks for Large-Scale Image Recognition.     | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [VGG19]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fvgg19.py)  | Very Deep Convolutional Networks for Large-Scale Image Recognition.     | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [ResNet]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fresnet_common.py)  | Deep Residual Learning for Image Recognition.     | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [ResNet50](https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fresnet50.py)  | Deep Residual Learning for Image Recognition.     | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [Nasnet](https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fnasnet.py)  | NASNet refers to Neural Architecture Search Network, a family of models that were designed automatically by learning the model architectures directly on the dataset of interest.    | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [MobileNet]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fmobilenet.py)  | MobileNet v1 models for Keras.  | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [MobileNet V2]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fmobilenet_v2.py)  | MobileNet v2 models for Keras.  | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [MobileNet V3]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fmobilenet_v3.py)  | MobileNet v3 models for Keras.  | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [efficientnet]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fefficientnet.py)  | Rethinking Model Scaling for Convolutional Neural Networks.  | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [Image analogies]( https:\u002F\u002Fgithub.com\u002Fawentzonline\u002Fimage-analogies)  | Generate image analogies using neural matching and blending.     | `Keras`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fawentzonline\u002Fimage-analogies\u002Fmaster\u002FLICENSE.txt )\n| [Popular Image Segmentation Models]( https:\u002F\u002Fgithub.com\u002Fdivamgupta\u002Fimage-segmentation-keras)  | Implementation of Segnet, FCN, UNet and other models in Keras.     | `Keras`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fdivamgupta\u002Fimage-segmentation-keras\u002Fmaster\u002FLICENSE )\n| [Ultrasound nerve segmentation]( https:\u002F\u002Fgithub.com\u002Fjocicmarko\u002Fultrasound-nerve-segmentation)  | This tutorial shows how to use Keras library to build deep neural network for ultrasound image nerve segmentation.     | `Keras`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjocicmarko\u002Fultrasound-nerve-segmentation\u002Fmaster\u002FLICENSE.md )\n| [DeepMask object segmentation]( https:\u002F\u002Fgithub.com\u002Fabbypa\u002FNNProject_DeepMask)  | This is a Keras-based Python implementation of DeepMask- a complex deep neural network for learning object segmentation masks.     | `Keras`| Not Found\n| [Monolingual and Multilingual Image Captioning]( https:\u002F\u002Fgithub.com\u002Felliottd\u002FGroundedTranslation)  | This is the source code that accompanies Multilingual Image Description with Neural Sequence Models.     | `Keras`| [BSD-3-Clause License]( https:\u002F\u002Fraw.githubusercontent.com\u002Felliottd\u002FGroundedTranslation\u002Fmaster\u002FLICENSE )\n| [pix2pix]( https:\u002F\u002Fgithub.com\u002Ftdeboissiere\u002FDeepLearningImplementations\u002Ftree\u002Fmaster\u002Fpix2pix)  | Keras implementation of Image-to-Image Translation with Conditional Adversarial Networks by Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, Alexei A.    | `Keras`| Not Found\n| [Colorful Image colorization]( https:\u002F\u002Fgithub.com\u002Ftdeboissiere\u002FDeepLearningImplementations\u002Ftree\u002Fmaster\u002FColorful)  | B&W to color.   | `Keras`| Not Found\n| [CycleGAN]( https:\u002F\u002Fgithub.com\u002Feriklindernoren\u002FKeras-GAN\u002Fblob\u002Fmaster\u002Fcyclegan\u002Fcyclegan.py)  | Implementation of _Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks_.    | `Keras`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Feriklindernoren\u002FKeras-GAN\u002Fmaster\u002FLICENSE )\n| [DualGAN](https:\u002F\u002Fgithub.com\u002Feriklindernoren\u002FKeras-GAN\u002Fblob\u002Fmaster\u002Fdualgan\u002Fdualgan.py)  | Implementation of _DualGAN: Unsupervised Dual Learning for Image-to-Image Translation_.   | `Keras`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Feriklindernoren\u002FKeras-GAN\u002Fmaster\u002FLICENSE )\n| [Super-Resolution GAN]( https:\u002F\u002Fgithub.com\u002Feriklindernoren\u002FKeras-GAN\u002Fblob\u002Fmaster\u002Fsrgan\u002Fsrgan.py)  | Implementation of _Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network_.   | `Keras`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Feriklindernoren\u002FKeras-GAN\u002Fmaster\u002FLICENSE )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ Back To Top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n### PyTorch \u003Ca name=\"pytorch\"\u002F>\n\n| Model Name | Description | Framework | License |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n|[detectron2](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetectron2) | Detectron2 is Facebook AI Research's next generation software system that implements state-of-the-art object detection algorithms | `PyTorch` | [Apache License 2.0](https:\u002F\u002Fraw.githubusercontent.com\u002Ffacebookresearch\u002Fdetectron2\u002Fmaster\u002FLICENSE) \n| [FastPhotoStyle]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FFastPhotoStyle)  | A Closed-form Solution to Photorealistic Image Stylization.   | `PyTorch`| [Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public Licens]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002FFastPhotoStyle\u002Fmaster\u002FLICENSE.md )\n| [pytorch-CycleGAN-and-pix2pix]( https:\u002F\u002Fgithub.com\u002Fjunyanz\u002Fpytorch-CycleGAN-and-pix2pix)  | A Closed-form Solution to Photorealistic Image Stylization.   | `PyTorch`| [BSD License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjunyanz\u002Fpytorch-CycleGAN-and-pix2pix\u002Fmaster\u002FLICENSE )\n| [maskrcnn-benchmark]( https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fmaskrcnn-benchmark)  | Fast, modular reference implementation of Instance Segmentation and Object Detection algorithms in PyTorch.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ffacebookresearch\u002Fmaskrcnn-benchmark\u002Fmaster\u002FLICENSE )\n| [deep-image-prior]( https:\u002F\u002Fgithub.com\u002FDmitryUlyanov\u002Fdeep-image-prior)  | Image restoration with neural networks but without learning.   | `PyTorch`| [Apache License 2.0]( https:\u002F\u002Fraw.githubusercontent.com\u002FDmitryUlyanov\u002Fdeep-image-prior\u002Fmaster\u002FLICENSE )\n| [StarGAN]( https:\u002F\u002Fgithub.com\u002Fyunjey\u002FStarGAN)  | StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fyunjey\u002FStarGAN\u002Fmaster\u002FLICENSE )\n| [faster-rcnn.pytorch]( https:\u002F\u002Fgithub.com\u002Fjwyang\u002Ffaster-rcnn.pytorch)  | This project is a faster faster R-CNN implementation, aimed to accelerating the training of faster R-CNN object detection models.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjwyang\u002Ffaster-rcnn.pytorch\u002Fmaster\u002FLICENSE )\n| [pix2pixHD]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fpix2pixHD)  | Synthesizing and manipulating 2048x1024 images with conditional GANs.  | `PyTorch`| [BSD License]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fpix2pixHD\u002Fmaster\u002FLICENSE.txt )\n| [Augmentor]( https:\u002F\u002Fgithub.com\u002Fmdbloice\u002FAugmentor)  | Image augmentation library in Python for machine learning.  | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmdbloice\u002FAugmentor\u002Fmaster\u002FLICENSE.md )\n| [albumentations]( https:\u002F\u002Fgithub.com\u002Falbumentations-team\u002Falbumentations)  | Fast image augmentation library.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Falbumentations-team\u002Falbumentations\u002Fmaster\u002FLICENSE )\n| [Deep Video Analytics]( https:\u002F\u002Fgithub.com\u002FAKSHAYUBHAT\u002FDeepVideoAnalytics)  | Deep Video Analytics is a platform for indexing and extracting information from videos and images   | `PyTorch`| [Custom]( https:\u002F\u002Fraw.githubusercontent.com\u002FAKSHAYUBHAT\u002FDeepVideoAnalytics\u002Fmaster\u002FLICENSE )\n| [semantic-segmentation-pytorch]( https:\u002F\u002Fgithub.com\u002FCSAILVision\u002Fsemantic-segmentation-pytorch)  | Pytorch implementation for Semantic Segmentation\u002FScene Parsing on MIT ADE20K dataset.   | `PyTorch`| [BSD 3-Clause License]( https:\u002F\u002Fraw.githubusercontent.com\u002FCSAILVision\u002Fsemantic-segmentation-pytorch\u002Fmaster\u002FLICENSE )\n| [An End-to-End Trainable Neural Network for Image-based Sequence Recognition]( https:\u002F\u002Fgithub.com\u002Fbgshih\u002Fcrnn)  | This software implements the Convolutional Recurrent Neural Network (CRNN), a combination of CNN, RNN and CTC loss for image-based sequence recognition tasks, such as scene text recognition and OCR.   | `PyTorch`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fbgshih\u002Fcrnn\u002Fmaster\u002FLICENSE )\n| [UNIT]( https:\u002F\u002Fgithub.com\u002Fmingyuliutw\u002FUNIT)  | PyTorch Implementation of our Coupled VAE-GAN algorithm for Unsupervised Image-to-Image Translation.   | `PyTorch`| [Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmingyuliutw\u002FUNIT\u002Fmaster\u002FLICENSE.md )\n| [Neural Sequence labeling model]( https:\u002F\u002Fgithub.com\u002Fjiesutd\u002FNCRFpp)  | Sequence labeling models are quite popular in many NLP tasks, such as Named Entity Recognition (NER), part-of-speech (POS) tagging and word segmentation.   | `PyTorch`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjiesutd\u002FNCRFpp\u002Fmaster\u002FLICENCE )\n| [faster rcnn]( https:\u002F\u002Fgithub.com\u002Flongcw\u002Ffaster_rcnn_pytorch)  | This is a PyTorch implementation of Faster RCNN. This project is mainly based on py-faster-rcnn and TFFRCNN. For details about R-CNN please refer to the paper Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks by Shaoqing Ren, Kaiming He, Ross Girshick, Jian Sun.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Flongcw\u002Ffaster_rcnn_pytorch\u002Fmaster\u002FLICENSE )\n| [pytorch-semantic-segmentation]( https:\u002F\u002Fgithub.com\u002FZijunDeng\u002Fpytorch-semantic-segmentation)  | PyTorch for Semantic Segmentation.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002FZijunDeng\u002Fpytorch-semantic-segmentation\u002Fmaster\u002FLICENSE )\n| [EDSR-PyTorch]( https:\u002F\u002Fgithub.com\u002Fthstkdgus35\u002FEDSR-PyTorch)  | PyTorch version of the paper 'Enhanced Deep Residual Networks for Single Image Super-Resolution'.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fthstkdgus35\u002FEDSR-PyTorch\u002Fmaster\u002FLICENSE )\n| [image-classification-mobile]( https:\u002F\u002Fgithub.com\u002Fosmr\u002Fimgclsmob)  | Collection of classification models pretrained on the ImageNet-1K.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fosmr\u002Fimgclsmob\u002Fmaster\u002FLICENSE )\n| [FaderNetworks]( https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FFaderNetworks)  | Fader Networks: Manipulating Images by Sliding Attributes - NIPS 2017.   | `PyTorch`| [Creative Commons Attribution-NonCommercial 4.0 International Public License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ffacebookresearch\u002FFaderNetworks\u002Fmaster\u002FLICENSE )\n| [neuraltalk2-pytorch]( https:\u002F\u002Fgithub.com\u002Fruotianluo\u002FImageCaptioning.pytorch)  | Image captioning model in pytorch (finetunable cnn in branch with_finetune).   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fruotianluo\u002FImageCaptioning.pytorch\u002Fmaster\u002FLICENSE )\n| [RandWireNN]( https:\u002F\u002Fgithub.com\u002Fseungwonpark\u002FRandWireNN)  | Implementation of: \"Exploring Randomly Wired Neural Networks for Image Recognition\".   | `PyTorch`| Not Found\n| [stackGAN-v2]( https:\u002F\u002Fgithub.com\u002Fhanzhanggit\u002FStackGAN-v2)  |Pytorch implementation for reproducing StackGAN_v2 results in the paper StackGAN++.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fhanzhanggit\u002FStackGAN-v2\u002Fmaster\u002FLICENSE )\n| [Detectron models for Object Detection]( https:\u002F\u002Fgithub.com\u002Fignacio-rocco\u002Fdetectorch)  | This code allows to use some of the Detectron models for object detection from Facebook AI Research with PyTorch.   | `PyTorch`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fignacio-rocco\u002Fdetectorch\u002Fmaster\u002FLICENSE )\n| [DEXTR-PyTorch]( https:\u002F\u002Fgithub.com\u002Fscaelles\u002FDEXTR-PyTorch)  | This paper explores the use of extreme points in an object (left-most, right-most, top, bottom pixels) as input to obtain precise object segmentation for images and videos.   | `PyTorch`| [GNU GENERAL PUBLIC LICENSE]( https:\u002F\u002Fraw.githubusercontent.com\u002Fscaelles\u002FDEXTR-PyTorch\u002Fmaster\u002FLICENSE )\n| [pointnet.pytorch]( https:\u002F\u002Fgithub.com\u002Ffxia22\u002Fpointnet.pytorch)  | Pytorch implementation for \"PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Ffxia22\u002Fpointnet.pytorch\u002Fmaster\u002FLICENSE )\n| [self-critical.pytorch]( https:\u002F\u002Fgithub.com\u002Fruotianluo\u002Fself-critical.pytorch) | This repository includes the unofficial implementation Self-critical Sequence Training for Image Captioning and Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fruotianluo\u002Fself-critical.pytorch\u002Fmaster\u002FLICENSE )\n| [vnet.pytorch]( https:\u002F\u002Fgithub.com\u002Fmattmacy\u002Fvnet.pytorch)  | A Pytorch implementation for V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation.   | `PyTorch`| [BSD 3-Clause License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmattmacy\u002Fvnet.pytorch\u002Fmaster\u002FLICENSE )\n| [piwise]( https:\u002F\u002Fgithub.com\u002Fbodokaiser\u002Fpiwise)  | Pixel-wise segmentation on VOC2012 dataset using pytorch.   | `PyTorch`| [BSD 3-Clause License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fbodokaiser\u002Fpiwise\u002Fmaster\u002FLICENSE.md )\n| [pspnet-pytorch]( https:\u002F\u002Fgithub.com\u002FLextal\u002Fpspnet-pytorch)  | PyTorch implementation of PSPNet segmentation network.   | `PyTorch`| Not Found\n| [pytorch-SRResNet]( https:\u002F\u002Fgithub.com\u002Ftwtygqyy\u002Fpytorch-SRResNet)  | Pytorch implementation for Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network.   | `PyTorch`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftwtygqyy\u002Fpytorch-SRResNet\u002Fmaster\u002FLICENSE )\n| [PNASNet.pytorch]( https:\u002F\u002Fgithub.com\u002Fchenxi116\u002FPNASNet.pytorch)  | PyTorch implementation of PNASNet-5 on ImageNet.   | `PyTorch`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fchenxi116\u002FPNASNet.pytorch\u002Fmaster\u002FLICENSE )\n| [img_classification_pk_pytorch]( https:\u002F\u002Fgithub.com\u002Ffelixgwu\u002Fimg_classification_pk_pytorch)  | Quickly comparing your image classification models with the state-of-the-art models.   | `PyTorch`| Not Found\n| [Deep Neural Networks are Easily Fooled]( https:\u002F\u002Fgithub.com\u002Futkuozbulak\u002Fpytorch-cnn-adversarial-attacks)  | High Confidence Predictions for Unrecognizable Images.   | `PyTorch`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Futkuozbulak\u002Fpytorch-cnn-adversarial-attacks\u002Fmaster\u002FLICENSE )\n| [pix2pix-pytorch]( https:\u002F\u002Fgithub.com\u002Fmrzhu-cool\u002Fpix2pix-pytorch)  | PyTorch implementation of \"Image-to-Image Translation Using Conditional Adversarial Networks\".   | `PyTorch`| Not Found\n| [NVIDIA\u002Fsemantic-segmentation]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fsemantic-segmentation)  | A PyTorch Implementation of Improving Semantic Segmentation via Video Propagation and Label Relaxation, In CVPR2019.   | `PyTorch`| [CC BY-NC-SA 4.0 license]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fsemantic-segmentation\u002Fmaster\u002FLICENSE )\n| [Neural-IMage-Assessment]( https:\u002F\u002Fgithub.com\u002Fkentsyx\u002FNeural-IMage-Assessment)  | A PyTorch Implementation of Neural IMage Assessment.   | `PyTorch`| Not Found\n| [torchxrayvision](https:\u002F\u002Fgithub.com\u002Fmlmed\u002Ftorchxrayvision) | Pretrained models for chest X-ray (CXR) pathology predictions. Medical, Healthcare, Radiology  | `PyTorch` | [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmlmed\u002Ftorchxrayvision\u002Fmaster\u002FLICENSE ) |\n| [pytorch-image-models](https:\u002F\u002Fgithub.com\u002Frwightman\u002Fpytorch-image-models) | PyTorch image models, scripts, pretrained weights -- (SE)ResNet\u002FResNeXT, DPN, EfficientNet, MixNet, MobileNet-V3\u002FV2, MNASNet, Single-Path NAS, FBNet, and more  | `PyTorch` | [Apache License 2.0]( https:\u002F\u002Fgithub.com\u002Frwightman\u002Fpytorch-image-models\u002Fblob\u002Fmaster\u002FLICENSE ) |\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ Back To Top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n### Caffe \u003Ca name=\"caffe\"\u002F>\n\n| Model Name | Description | Framework | License |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [OpenPose]( https:\u002F\u002Fgithub.com\u002FCMU-Perceptual-Computing-Lab\u002Fopenpose)  | OpenPose represents the first real-time multi-person system to jointly detect human body, hand, and facial keypoints (in total 130 keypoints) on single images.   | `Caffe`| [Custom]( https:\u002F\u002Fraw.githubusercontent.com\u002FCMU-Perceptual-Computing-Lab\u002Fopenpose\u002Fmaster\u002FLICENSE )\n| [Fully Convolutional Networks for Semantic Segmentation]( https:\u002F\u002Fgithub.com\u002Fshelhamer\u002Ffcn.berkeleyvision.org)  | Fully Convolutional Models for Semantic Segmentation.   | `Caffe`| Not Found\n| [Colorful Image Colorization]( https:\u002F\u002Fgithub.com\u002Frichzhang\u002Fcolorization)  | Colorful Image Colorization.   | `Caffe`| [BSD-2-Clause License]( https:\u002F\u002Fraw.githubusercontent.com\u002Frichzhang\u002Fcolorization\u002Fmaster\u002FLICENSE )\n| [R-FCN]( https:\u002F\u002Fgithub.com\u002FYuwenXiong\u002Fpy-R-FCN)  | R-FCN: Object Detection via Region-based Fully Convolutional Networks.   | `Caffe`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002FYuwenXiong\u002Fpy-R-FCN\u002Fmaster\u002FLICENSE )\n| [cnn-vis]( https:\u002F\u002Fgithub.com\u002Fjcjohnson\u002Fcnn-vis)  |Inspired by Google's recent Inceptionism blog post, cnn-vis is an open-source tool that lets you use convolutional neural networks to generate images.   | `Caffe`| [The MIT License (MIT)]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjcjohnson\u002Fcnn-vis\u002Fmaster\u002FLICENSE )\n| [DeconvNet]( https:\u002F\u002Fgithub.com\u002FHyeonwooNoh\u002FDeconvNet)  | Learning Deconvolution Network for Semantic Segmentation.   | `Caffe`| [Custom]( https:\u002F\u002Fraw.githubusercontent.com\u002FHyeonwooNoh\u002FDeconvNet\u002Fmaster\u002FLICENSE )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ Back To Top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n### MXNet \u003Ca name=\"mxnet\"\u002F>\n\n| Model Name | Description | Framework | License |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [Faster RCNN]( https:\u002F\u002Fgithub.com\u002Fijkguo\u002Fmx-rcnn)  | Region Proposal Network solves object detection as a regression problem.   | `MXNet`| [Apache License, Version 2.0]( https:\u002F\u002Fraw.githubusercontent.com\u002Fijkguo\u002Fmx-rcnn\u002Fmaster\u002FLICENSE )\n| [SSD]( https:\u002F\u002Fgithub.com\u002Fzhreshold\u002Fmxnet-ssd)  | SSD is an unified framework for object detection with a single network.   | `MXNet`| [MIT License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fzhreshold\u002Fmxnet-ssd\u002Fmaster\u002FLICENSE )\n| [Faster RCNN+Focal Loss]( https:\u002F\u002Fgithub.com\u002Funsky\u002Ffocal-loss)  | The code is unofficial version for focal loss for Dense Object Detection.   | `MXNet`| Not Found\n| [CNN-LSTM-CTC]( https:\u002F\u002Fgithub.com\u002Foyxhust\u002FCNN-LSTM-CTC-text-recognition)  |I realize three different models for text recognition, and all of them consist of CTC loss layer to realize no segmentation for text images.   | `MXNet`| Not Found\n| [Faster_RCNN_for_DOTA]( https:\u002F\u002Fgithub.com\u002Fjessemelpolio\u002FFaster_RCNN_for_DOTA)  | This is the official repo of paper _DOTA: A Large-scale Dataset for Object Detection in Aerial Images_.  | `MXNet`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjessemelpolio\u002FFaster_RCNN_for_DOTA\u002Fmaster\u002FLICENSE )\n| [RetinaNet]( https:\u002F\u002Fgithub.com\u002Funsky\u002FRetinaNet)  | Focal loss for Dense Object Detection.   | `MXNet`| Not Found\n| [MobileNetV2]( https:\u002F\u002Fgithub.com\u002Fliangfu\u002Fmxnet-mobilenet-v2)  | This is a MXNet implementation of MobileNetV2 architecture as described in the paper _Inverted Residuals and Linear Bottlenecks: Mobile Networks for Classification, Detection and Segmentation_.   | `MXNet`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fliangfu\u002Fmxnet-mobilenet-v2\u002Fmaster\u002FLICENSE )\n| [neuron-selectivity-transfer]( https:\u002F\u002Fgithub.com\u002FTuSimple\u002Fneuron-selectivity-transfer)  | This code is a re-implementation of the imagenet classification experiments in the paper _Like What You Like: Knowledge Distill via Neuron Selectivity Transfer_.   | `MXNet`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002FTuSimple\u002Fneuron-selectivity-transfer\u002Fmaster\u002FLICENSE )\n| [MobileNetV2]( https:\u002F\u002Fgithub.com\u002Fchinakook\u002FMobileNetV2.mxnet)  | This is a Gluon implementation of MobileNetV2 architecture as described in the paper _Inverted Residuals and Linear Bottlenecks: Mobile Networks for Classification, Detection and Segmentation_.   | `MXNet`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002Fchinakook\u002FMobileNetV2.mxnet\u002Fmaster\u002FLICENSE )\n| [sparse-structure-selection]( https:\u002F\u002Fgithub.com\u002FTuSimple\u002Fsparse-structure-selection)  | This code is a re-implementation of the imagenet classification experiments in the paper _Data-Driven Sparse Structure Selection for Deep Neural Networks_.   | `MXNet`| [Apache License]( https:\u002F\u002Fraw.githubusercontent.com\u002FTuSimple\u002Fsparse-structure-selection\u002Fmaster\u002FLICENSE )\n| [FastPhotoStyle]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FFastPhotoStyle)  | A Closed-form Solution to Photorealistic Image Stylization.   | `MXNet`| [Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002FFastPhotoStyle\u002Fmaster\u002FLICENSE.md )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ Back To Top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n## Contributions\nYour contributions are always welcome!!\nPlease have a look at `contributing.md`\n\n## License\n\n[MIT License](LICENSE)\n","![维护](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FMaintained%3F-YES-green.svg)\n![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FRelease-PROD-yellow.svg)\n![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLanguages-MULTI-blue.svg)\n![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-lightgrey.svg)\n\n# 计算机视觉预训练模型\n\n![CV标志](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fbalavenkatesh3322_CV-pretrained-model_readme_dd3d9bf50810.jpg)\n\n## 什么是预训练模型？\n预训练模型是由他人为解决类似问题而创建的模型。与其从头开始构建一个解决类似问题的模型，不如将已在其他问题上训练好的模型作为起点。不过，预训练模型在你的应用场景中可能并不完全准确。\n\n例如，如果你想开发一辆自动驾驶汽车，你可以花费数年时间从零开始构建一个像样的图像识别算法；或者，你可以直接使用谷歌提供的Inception模型（一种预训练模型），该模型基于[ImageNet](http:\u002F\u002Fwww.image-net.org\u002F)数据集训练而成，能够识别图片中的内容。\n\n## 其他预训练模型\n* [NLP预训练模型](https:\u002F\u002Fgithub.com\u002Fbalavenkatesh3322\u002FNLP-pretrained-model)。\n* [音频与语音预训练模型](https:\u002F\u002Fgithub.com\u002Fbalavenkatesh3322\u002Faudio-pretrained-model)。\n\n## 模型部署库\n* [模型服务](https:\u002F\u002Fgithub.com\u002Fbalavenkatesh3322\u002Fmodel_deployment)\n\n### 框架\n\n* [TensorFlow](#tensorflow)\n* [Keras](#keras)\n* [PyTorch](#pytorch)\n* [Caffe](#caffe)\n* [MXNet](#mxnet)\n\n### 模型可视化\n你可以使用[Netron](https:\u002F\u002Fgithub.com\u002Flutzroeder\u002FNetron)查看每个模型的网络架构可视化图。\n\n![CV标志](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fbalavenkatesh3322_CV-pretrained-model_readme_76d47d615285.png)\n\n### TensorFlow \u003Ca name=\"tensorflow\"\u002F>\n\n| 模型名称 | 描述 | 框架 | 许可证 |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [目标检测]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fobject_detection)  | 在单张图像中定位并识别多个对象。| `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [Mask R-CNN]( https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)  | 该模型为图像中每个对象实例生成边界框和分割掩码。它基于特征金字塔网络（FPN）和ResNet101骨干网络。     | `TensorFlow`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmatterport\u002FMask_RCNN\u002Fmaster\u002FLICENSE )\n| [Faster-RCNN]( https:\u002F\u002Fgithub.com\u002Fsmallcorgi\u002FFaster-RCNN_TF)  | 这是Faster RCNN的实验性TensorFlow实现——一种带有区域建议网络的目标检测卷积神经网络。     | `TensorFlow`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fsmallcorgi\u002FFaster-RCNN_TF\u002Fmaster\u002FLICENSE )\n| [YOLO TensorFlow]( https:\u002F\u002Fgithub.com\u002Fgliese581gg\u002FYOLO_tensorflow)  | 这是YOLO：实时目标检测的TensorFlow实现。     | `TensorFlow`| [自定义]( https:\u002F\u002Fraw.githubusercontent.com\u002Fgliese581gg\u002FYOLO_tensorflow\u002Fmaster\u002FLICENSE )\n| [YOLO TensorFlow ++]( https:\u002F\u002Fgithub.com\u002Fthtrieu\u002Fdarkflow)  | “YOLO：实时目标检测”的TensorFlow实现，支持训练并在移动设备上实现实时运行。     | `TensorFlow`| [GNU通用公共许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fthtrieu\u002Fdarkflow\u002Fmaster\u002FLICENSE )\n| [MobileNet]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Fblob\u002Fmaster\u002Fresearch\u002Fslim\u002Fnets\u002Fmobilenet_v1.md)  | MobileNets在延迟、大小和精度之间进行权衡，同时与文献中流行的模型相比具有优势。     | `TensorFlow`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [DeepLab]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fdeeplab)  | 用于语义图像分割的深度标注。     | `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [Colornet]( https:\u002F\u002Fgithub.com\u002Fpavelgonchar\u002Fcolornet)  | 将灰度图像着色的神经网络。     | `TensorFlow`| 未找到\n| [SRGAN]( https:\u002F\u002Fgithub.com\u002Ftensorlayer\u002Fsrgan)  | 使用生成对抗网络实现照片级真实的单幅图像超分辨率。    | `TensorFlow`| 未找到\n| [DeepOSM]( https:\u002F\u002Fgithub.com\u002Ftrailbehind\u002FDeepOSM)  | 使用OpenStreetMap特征和卫星图像训练TensorFlow神经网络。     | `TensorFlow`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftrailbehind\u002FDeepOSM\u002Fmaster\u002FLICENSE )\n| [领域迁移网络]( https:\u002F\u002Fgithub.com\u002Fyunjey\u002Fdomain-transfer-network)  | 无监督跨域图像生成的实现。  | `TensorFlow`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fyunjey\u002Fdomain-transfer-network\u002Fmaster\u002FLICENSE )\n| [Show, Attend and Tell]( https:\u002F\u002Fgithub.com\u002Fyunjey\u002Fshow-attend-and-tell)  | 基于注意力机制的图像字幕生成器。     | `TensorFlow`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fyunjey\u002Fshow-attend-and-tell\u002Fmaster\u002FLICENSE )\n| [android-yolo]( https:\u002F\u002Fgithub.com\u002Fnatanielruiz\u002Fandroid-yolo)  | 使用YOLO网络和TensorFlow在Android设备上实现实时目标检测。    | `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fnatanielruiz\u002Fandroid-yolo\u002Fmaster\u002FLICENSE )\n| [DCSCN超分辨率]( https:\u002F\u002Fgithub.com\u002Fjiny2001\u002Fdcscn-super-resolutiont)  | 这是“通过带有跳跃连接和网络内网络的深度CNN实现快速准确的图像超分辨率”的TensorFlow实现，是一种基于深度学习的单幅图像超分辨率（SISR）模型。     | `TensorFlow`| 未找到\n| [GAN-CLS]( https:\u002F\u002Fgithub.com\u002Fzsdonghao\u002Ftext-to-image)  | 这是合成图像的实验性TensorFlow实现。     | `TensorFlow`| 未找到\n| [U-Net]( https:\u002F\u002Fgithub.com\u002Fzsdonghao\u002Fu-net-brain-tumor)  | 用于脑肿瘤分割。     | `TensorFlow`| 未找到\n| [改进的CycleGAN]( https:\u002F\u002Fgithub.com\u002Fluoxier\u002FCycleGAN_Tensorlayer)  | 无配对图像到图像的转换。     | `TensorFlow`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fluoxier\u002FCycleGAN_Tensorlayer\u002Fmaster\u002FLICENSE )\n| [Im2txt]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fim2txt)  | 用于图像字幕生成的图像到文本神经网络。     | `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [SLIM]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fslim)  | TF-Slim中的图像分类模型。     | `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [DELF]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fdelf)  | 用于图像匹配和检索的深度局部特征。     | `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [压缩]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fcompression)  | 使用预训练的残差GRU网络对图像进行压缩和解压缩。     | `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n| [AttentionOCR]( https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Ftree\u002Fmaster\u002Fresearch\u002Fattention_ocr)  | 用于从真实世界图像中提取文本的模型。     | `TensorFlow`| [Apache许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftensorflow\u002Fmodels\u002Fmaster\u002FLICENSE )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ 回到顶部\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n### Keras \u003Ca name=\"keras\"\u002F>\n\n| 模型名称 | 描述 | 框架 | 许可证 |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [Mask R-CNN]( https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN)  | 该模型为图像中每个对象实例生成边界框和分割掩码。它基于特征金字塔网络（FPN）和ResNet101骨干网络。| `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmatterport\u002FMask_RCNN\u002Fmaster\u002FLICENSE )\n| [VGG16]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fvgg16.py)  | 用于大规模图像识别的非常深的卷积神经网络。     | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [VGG19]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fvgg19.py)  | 用于大规模图像识别的非常深的卷积神经网络。     | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [ResNet]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fresnet_common.py)  | 用于图像识别的深度残差学习。     | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [ResNet50](https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fresnet50.py)  | 用于图像识别的深度残差学习。     | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [Nasnet](https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fnasnet.py)  | NASNet指的是神经架构搜索网络，这是一系列通过直接在感兴趣的数据集上学习模型架构而自动设计出来的模型。    | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [MobileNet]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fmobilenet.py)  | 适用于Keras的MobileNet v1模型。  | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [MobileNet V2]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fmobilenet_v2.py)  | 适用于Keras的MobileNet v2模型。  | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [MobileNet V3]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fmobilenet_v3.py)  | 适用于Keras的MobileNet v3模型。  | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [efficientnet]( https:\u002F\u002Fgithub.com\u002Fkeras-team\u002Fkeras-applications\u002Fblob\u002Fmaster\u002Fkeras_applications\u002Fefficientnet.py)  | 对卷积神经网络的模型缩放进行重新思考。  | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fkeras-team\u002Fkeras-applications\u002Fmaster\u002FLICENSE )\n| [图像类比]( https:\u002F\u002Fgithub.com\u002Fawentzonline\u002Fimage-analogies)  | 使用神经匹配和混合生成图像类比。     | `Keras`| [MIT许可证（MIT）]( https:\u002F\u002Fraw.githubusercontent.com\u002Fawentzonline\u002Fimage-analogies\u002Fmaster\u002FLICENSE.txt )\n| [流行的图像分割模型]( https:\u002F\u002Fgithub.com\u002Fdivamgupta\u002Fimage-segmentation-keras)  | 在Keras中实现Segnet、FCN、UNet等模型。     | `Keras`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fdivamgupta\u002Fimage-segmentation-keras\u002Fmaster\u002FLICENSE )\n| [超声神经分割]( https:\u002F\u002Fgithub.com\u002Fjocicmarko\u002Fultrasound-nerve-segmentation)  | 本教程展示了如何使用Keras库构建用于超声图像神经分割的深度神经网络。     | `Keras`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjocicmarko\u002Fultrasound-nerve-segmentation\u002Fmaster\u002FLICENSE.md )\n| [DeepMask对象分割]( https:\u002F\u002Fgithub.com\u002Fabbypa\u002FNNProject_DeepMask)  | 这是基于Keras的Python实现，用于学习对象分割掩码的复杂深度神经网络DeepMask。     | `Keras`| 未找到\n| [单语和多语图像描述]( https:\u002F\u002Fgithub.com\u002Felliottd\u002FGroundedTranslation)  | 这是伴随《使用神经序列模型的多语图像描述》一书的源代码。     | `Keras`| [BSD-3-Clause许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Felliottd\u002FGroundedTranslation\u002Fmaster\u002FLICENSE )\n| [pix2pix]( https:\u002F\u002Fgithub.com\u002Ftdeboissiere\u002FDeepLearningImplementations\u002Ftree\u002Fmaster\u002Fpix2pix)  | 由Phillip Isola、Jun-Yan Zhu、Tinghui Zhou、Alexei A.等人提出的条件对抗网络图像到图像转换的Keras实现。    | `Keras`| 未找到\n| [彩色图像着色]( https:\u002F\u002Fgithub.com\u002Ftdeboissiere\u002FDeepLearningImplementations\u002Ftree\u002Fmaster\u002FColorful)  | 黑白转彩色。   | `Keras`| 未找到\n| [CycleGAN]( https:\u002F\u002Fgithub.com\u002Feriklindernoren\u002FKeras-GAN\u002Fblob\u002Fmaster\u002Fcyclegan\u002Fcyclegan.py)  | 实现了“使用循环一致性对抗网络的非配对图像到图像转换”。    | `Keras`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Feriklindernoren\u002FKeras-GAN\u002Fmaster\u002FLICENSE )\n| [DualGAN](https:\u002F\u002Fgithub.com\u002Feriklindernoren\u002FKeras-GAN\u002Fblob\u002Fmaster\u002Fdualgan\u002Fdualgan.py)  | 实现了“DualGAN：用于图像到图像转换的无监督双学习”。   | `Keras`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Feriklindernoren\u002FKeras-GAN\u002Fmaster\u002FLICENSE )\n| [超分辨率GAN]( https:\u002F\u002Fgithub.com\u002Feriklindernoren\u002FKeras-GAN\u002Fblob\u002Fmaster\u002Fsrgan\u002Fsrgan.py)  | 实现了“使用生成对抗网络进行照片级真实感单张图像超分辨率”。   | `Keras`| [MIT许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Feriklindernoren\u002FKeras-GAN\u002Fmaster\u002FLICENSE )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ 返回顶部\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n### PyTorch \u003Ca name=\"pytorch\"\u002F>\n\n| 模型名称 | 描述 | 框架 | 许可证 |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n|[detectron2](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdetectron2) | Detectron2 是 Facebook AI Research 的下一代软件系统，实现了最先进的目标检测算法 | `PyTorch` | [Apache License 2.0](https:\u002F\u002Fraw.githubusercontent.com\u002Ffacebookresearch\u002Fdetectron2\u002Fmaster\u002FLICENSE) \n| [FastPhotoStyle]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FFastPhotoStyle)  | 一种用于照片级真实感图像风格化的闭式解。   | `PyTorch`| [知识共享署名-非商业性使用-相同方式共享 4.0 国际许可协议]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002FFastPhotoStyle\u002Fmaster\u002FLICENSE.md )\n| [pytorch-CycleGAN-and-pix2pix]( https:\u002F\u002Fgithub.com\u002Fjunyanz\u002Fpytorch-CycleGAN-and-pix2pix)  | 一种用于照片级真实感图像风格化的闭式解。   | `PyTorch`| [BSD 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjunyanz\u002Fpytorch-CycleGAN-and-pix2pix\u002Fmaster\u002FLICENSE )\n| [maskrcnn-benchmark]( https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fmaskrcnn-benchmark)  | 在 PyTorch 中快速、模块化的实例分割和目标检测算法参考实现。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ffacebookresearch\u002Fmaskrcnn-benchmark\u002Fmaster\u002FLICENSE )\n| [deep-image-prior]( https:\u002F\u002Fgithub.com\u002FDmitryUlyanov\u002Fdeep-image-prior)  | 使用神经网络进行图像恢复，但无需学习。   | `PyTorch`| [Apache License 2.0]( https:\u002F\u002Fraw.githubusercontent.com\u002FDmitryUlyanov\u002Fdeep-image-prior\u002Fmaster\u002FLICENSE )\n| [StarGAN]( https:\u002F\u002Fgithub.com\u002Fyunjey\u002FStarGAN)  | StarGAN：用于多领域图像到图像转换的统一生成对抗网络。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fyunjey\u002FStarGAN\u002Fmaster\u002FLICENSE )\n| [faster-rcnn.pytorch]( https:\u002F\u002Fgithub.com\u002Fjwyang\u002Ffaster-rcnn.pytorch)  | 该项目是一个更快的 Faster R-CNN 实现，旨在加速 Faster R-CNN 目标检测模型的训练。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjwyang\u002Ffaster-rcnn.pytorch\u002Fmaster\u002FLICENSE )\n| [pix2pixHD]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fpix2pixHD)  | 使用条件 GAN 合成和操作 2048x1024 分辨率的图像。   | `PyTorch`| [BSD 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fpix2pixHD\u002Fmaster\u002FLICENSE.txt )\n| [Augmentor]( https:\u002F\u002Fgithub.com\u002Fmdbloice\u002FAugmentor)  | 用于机器学习的 Python 图像增强库。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmdbloice\u002FAugmentor\u002Fmaster\u002FLICENSE.md )\n| [albumentations]( https:\u002F\u002Fgithub.com\u002Falbumentations-team\u002Falbumentations)  | 快速图像增强库。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Falbumentations-team\u002Falbumentations\u002Fmaster\u002FLICENSE )\n| [Deep Video Analytics]( https:\u002F\u002Fgithub.com\u002FAKSHAYUBHAT\u002FDeepVideoAnalytics)  | Deep Video Analytics 是一个用于对视频和图像进行索引和信息提取的平台   | `PyTorch`| [自定义许可]( https:\u002F\u002Fraw.githubusercontent.com\u002FAKSHAYUBHAT\u002FDeepVideoAnalytics\u002Fmaster\u002FLICENSE )\n| [semantic-segmentation-pytorch]( https:\u002F\u002Fgithub.com\u002FCSAILVision\u002Fsemantic-segmentation-pytorch)  | MIT ADE20K 数据集上的语义分割\u002F场景解析的 PyTorch 实现。   | `PyTorch`| [BSD 3-Clause 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002FCSAILVision\u002Fsemantic-segmentation-pytorch\u002Fmaster\u002FLICENSE )\n| [基于图像序列识别的端到端可训练神经网络]( https:\u002F\u002Fgithub.com\u002Fbgshih\u002Fcrnn)  | 该软件实现了卷积循环神经网络 (CRNN)，它是 CNN、RNN 和 CTC 损失的结合体，适用于基于图像的序列识别任务，如场景文本识别和 OCR。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fbgshih\u002Fcrnn\u002Fmaster\u002FLICENSE )\n| [UNIT]( https:\u002F\u002Fgithub.com\u002Fmingyuliutw\u002FUNIT)  | 我们用于无监督图像到图像转换的耦合 VAE-GAN 算法的 PyTorch 实现。   | `PyTorch`| [知识共享署名-非商业性使用-相同方式共享 4.0 国际公共许可]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmingyuliutw\u002FUNIT\u002Fmaster\u002FLICENSE.md )\n| [神经序列标注模型]( https:\u002F\u002Fgithub.com\u002Fjiesutd\u002FNCRFpp)  | 序列标注模型在许多 NLP 任务中非常流行，例如命名实体识别 (NER)、词性标注 (POS) 和分词。   | `PyTorch`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjiesutd\u002FNCRFpp\u002Fmaster\u002FLICENCE )\n| [faster rcnn]( https:\u002F\u002Fgithub.com\u002Flongcw\u002Ffaster_rcnn_pytorch)  | 这是 Faster RCNN 的 PyTorch 实现。该项目主要基于 py-faster-rcnn 和 TFFRCNN。有关 R-CNN 的详细信息，请参阅 Shaoqing Ren、Kaiming He、Ross Girshick 和 Jian Sun 的论文《Faster R-CNN：通过区域建议网络实现实时目标检测》。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Flongcw\u002Ffaster_rcnn_pytorch\u002Fmaster\u002FLICENSE )\n| [pytorch-semantic-segmentation]( https:\u002F\u002Fgithub.com\u002FZijunDeng\u002Fpytorch-semantic-segmentation)  | 用于语义分割的 PyTorch。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002FZijunDeng\u002Fpytorch-semantic-segmentation\u002Fmaster\u002FLICENSE )\n| [EDSR-PyTorch]( https:\u002F\u002Fgithub.com\u002Fthstkdgus35\u002FEDSR-PyTorch)  | 论文《用于单张图像超分辨率的增强深度残差网络》的 PyTorch 版本。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fthstkdgus35\u002FEDSR-PyTorch\u002Fmaster\u002FLICENSE )\n| [image-classification-mobile]( https:\u002F\u002Fgithub.com\u002Fosmr\u002Fimgclsmob)  | ImageNet-1K 上预训练的分类模型集合。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fosmr\u002Fimgclsmob\u002Fmaster\u002FLICENSE )\n| [FaderNetworks]( https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FFaderNetworks)  | Fader Networks：通过滑动属性操纵图像——NIPS 2017。   | `PyTorch`| [知识共享署名-非商业性使用 4.0 国际公共许可]( https:\u002F\u002Fraw.githubusercontent.com\u002Ffacebookresearch\u002FFaderNetworks\u002Fmaster\u002FLICENSE )\n| [neuraltalk2-pytorch]( https:\u002F\u002Fgithub.com\u002Fruotianluo\u002FImageCaptioning.pytorch)  | PyTorch 中的图像字幕模型（带有 finetune 分支的可微调 CNN）。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fruotianluo\u002FImageCaptioning.pytorch\u002Fmaster\u002FLICENSE )\n| [RandWireNN]( https:\u002F\u002Fgithub.com\u002Fseungwonpark\u002FRandWireNN)  | 实现：“探索随机连接的神经网络用于图像识别”。   | `PyTorch`| 未找到\n| [stackGAN-v2]( https:\u002F\u002Fgithub.com\u002Fhanzhanggit\u002FStackGAN-v2)  | PyTorch 实现，用于复现论文 StackGAN++ 中的 StackGAN_v2 结果。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fhanzhanggit\u002FStackGAN-v2\u002Fmaster\u002FLICENSE )\n| [Detectron 目标检测模型]( https:\u002F\u002Fgithub.com\u002Fignacio-rocco\u002Fdetectorch)  | 此代码允许使用来自 Facebook AI Research 的部分 Detectron 目标检测模型，并与 PyTorch 配合使用。   | `PyTorch`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fignacio-rocco\u002Fdetectorch\u002Fmaster\u002FLICENSE )\n| [DEXTR-PyTorch]( https:\u002F\u002Fgithub.com\u002Fscaelles\u002FDEXTR-PyTorch)  | 本文探讨了将物体的极端点（最左、最右、顶部、底部像素）作为输入，以获得精确的图像和视频对象分割。   | `PyTorch`| [GNU 通用公共许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fscaelles\u002FDEXTR-PyTorch\u002Fmaster\u002FLICENSE )\n| [pointnet.pytorch]( https:\u002F\u002Fgithub.com\u002Ffxia22\u002Fpointnet.pytorch)  | “PointNet：用于 3D 分类和分割的点云深度学习”的 PyTorch 实现。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ffxia22\u002Fpointnet.pytorch\u002Fmaster\u002FLICENSE )\n| [self-critical.pytorch]( https:\u002F\u002Fgithub.com\u002Fruotianluo\u002Fself-critical.pytorch) | 该仓库包含非官方实现“用于图像字幕的自我批判序列训练”以及“用于图像字幕和视觉问答的自下而上和自上而下的注意力”。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fruotianluo\u002Fself-critical.pytorch\u002Fmaster\u002FLICENSE )\n| [vnet.pytorch]( https:\u002F\u002Fgithub.com\u002Fmattmacy\u002Fvnet.pytorch)  | V-Net 的 PyTorch 实现：用于体积医学图像分割的全卷积神经网络。   | `PyTorch`| [BSD 3-Clause 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmattmacy\u002Fvnet.pytorch\u002Fmaster\u002FLICENSE )\n| [piwise]( https:\u002F\u002Fgithub.com\u002Fbodokaiser\u002Fpiwise)  | 使用 PyTorch 对 VOC2012 数据集进行逐像素分割。   | `PyTorch`| [BSD 3-Clause 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fbodokaiser\u002Fpiwise\u002Fmaster\u002FLICENSE.md )\n| [pspnet-pytorch]( https:\u002F\u002Fgithub.com\u002FLextal\u002Fpspnet-pytorch)  | PSPNet 分割网络的 PyTorch 实现。   | `PyTorch`| 未找到\n| [pytorch-SRResNet]( https:\u002F\u002Fgithub.com\u002Ftwtygqyy\u002Fpytorch-SRResNet)  | 使用生成对抗网络实现照片级真实感单张图像超分辨率的 PyTorch 实现。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Ftwtygqyy\u002Fpytorch-SRResNet\u002Fmaster\u002FLICENSE )\n| [PNASNet.pytorch]( https:\u002F\u002Fgithub.com\u002Fchenxi116\u002FPNASNet.pytorch)  | PNASNet-5 在 ImageNet 上的 PyTorch 实现。   | `PyTorch`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fchenxi116\u002FPNASNet.pytorch\u002Fmaster\u002FLICENSE )\n| [img_classification_pk_pytorch]( https:\u002F\u002Fgithub.com\u002Ffelixgwu\u002Fimg_classification_pk_pytorch)  | 快速比较您的图像分类模型与最先进模型。   | `PyTorch`| 未找到\n| [深度神经网络很容易被欺骗]( https:\u002F\u002Fgithub.com\u002Futkuozbulak\u002Fpytorch-cnn-adversarial-attacks)  | 对无法识别的图像做出高度自信的预测。   | `PyTorch`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Futkuozbulak\u002Fpytorch-cnn-adversarial-attacks\u002Fmaster\u002FLICENSE )\n| [pix2pix-pytorch]( https:\u002F\u002Fgithub.com\u002Fmrzhu-cool\u002Fpix2pix-pytorch)  | “使用条件对抗网络进行图像到图像转换”的 PyTorch 实现。   | `PyTorch`| 未找到\n| [NVIDIA\u002Fsemantic-segmentation]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fsemantic-segmentation)  | 在 CVPR2019 上提出的通过视频传播和标签松弛来改进语义分割的 PyTorch 实现。   | `PyTorch`| [CC BY-NC-SA 4.0 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fsemantic-segmentation\u002Fmaster\u002FLICENSE )\n| [Neural-IMage-Assessment]( https:\u002F\u002Fgithub.com\u002Fkentsyx\u002FNeural-IMage-Assessment)  | 神经图像评估的 PyTorch 实现。   | `PyTorch`| 未找到\n| [torchxrayvision](https:\u002F\u002Fgithub.com\u002Fmlmed\u002Ftorchxrayvision) | 用于胸部 X 光 (CXR) 病理预测的预训练模型。医疗、健康护理、放射学  | `PyTorch` | [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fmlmed\u002Ftorchxrayvision\u002Fmaster\u002FLICENSE ) |\n| [pytorch-image-models]( https:\u002F\u002Fgithub.com\u002Frwightman\u002Fpytorch-image-models) | PyTorch 图像模型、脚本、预训练权重——(SE)ResNet\u002FResNeXT、DPN、EfficientNet、MixNet、MobileNet-V3\u002FV2、MNASNet、Single-Path NAS、FBNet 等  | `PyTorch` | [Apache License 2.0]( https:\u002F\u002Fgithub.com\u002Frwightman\u002Fpytorch-image-models\u002Fblob\u002Fmaster\u002FLICENSE ) |\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ 返回顶部\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n\n### Caffe \u003Ca name=\"caffe\"\u002F>\n\n| 模型名称 | 描述 | 框架 | 许可证 |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [OpenPose]( https:\u002F\u002Fgithub.com\u002FCMU-Perceptual-Computing-Lab\u002Fopenpose)  | OpenPose 是首个能够在单张图像上同时检测人体、手部和面部关键点（共130个关键点）的实时多人系统。   | `Caffe`| [自定义]( https:\u002F\u002Fraw.githubusercontent.com\u002FCMU-Perceptual-Computing-Lab\u002Fopenpose\u002Fmaster\u002FLICENSE )\n| [用于语义分割的全卷积网络]( https:\u002F\u002Fgithub.com\u002Fshelhamer\u002Ffcn.berkeleyvision.org)  | 用于语义分割的全卷积模型。   | `Caffe`| 未找到\n| [彩色图像着色]( https:\u002F\u002Fgithub.com\u002Frichzhang\u002Fcolorization)  | 彩色图像着色。   | `Caffe`| [BSD-2-Clause 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Frichzhang\u002Fcolorization\u002Fmaster\u002FLICENSE )\n| [R-FCN]( https:\u002F\u002Fgithub.com\u002FYuwenXiong\u002Fpy-R-FCN)  | R-FCN：基于区域的全卷积网络目标检测。   | `Caffe`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002FYuwenXiong\u002Fpy-R-FCN\u002Fmaster\u002FLICENSE )\n| [cnn-vis]( https:\u002F\u002Fgithub.com\u002Fjcjohnson\u002Fcnn-vis)  | 受谷歌近期 Inceptionism 博客文章启发，cnn-vis 是一款开源工具，允许使用卷积神经网络生成图像。   | `Caffe`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjcjohnson\u002Fcnn-vis\u002Fmaster\u002FLICENSE )\n| [DeconvNet]( https:\u002F\u002Fgithub.com\u002FHyeonwooNoh\u002FDeconvNet)  | 学习用于语义分割的反卷积网络。   | `Caffe`| [自定义]( https:\u002F\u002Fraw.githubusercontent.com\u002FHyeonwooNoh\u002FDeconvNet\u002Fmaster\u002FLICENSE )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ 返回顶部\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n### MXNet \u003Ca name=\"mxnet\"\u002F>\n\n| 模型名称 | 描述 | 框架 | 许可证 |\n|   :---:      |     :---:      |     :---:     |     :---:     |\n| [Faster RCNN]( https:\u002F\u002Fgithub.com\u002Fijkguo\u002Fmx-rcnn)  | 区域建议网络将目标检测问题转化为回归问题。   | `MXNet`| [Apache 许可证，版本 2.0]( https:\u002F\u002Fraw.githubusercontent.com\u002Fijkguo\u002Fmx-rcnn\u002Fmaster\u002FLICENSE )\n| [SSD]( https:\u002F\u002Fgithub.com\u002Fzhreshold\u002Fmxnet-ssd)  | SSD 是一种使用单一网络进行目标检测的统一框架。   | `MXNet`| [MIT 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fzhreshold\u002Fmxnet-ssd\u002Fmaster\u002FLICENSE )\n| [Faster RCNN+焦点损失]( https:\u002F\u002Fgithub.com\u002Funsky\u002Ffocal-loss)  | 该代码是针对密集目标检测的焦点损失的非官方版本。   | `MXNet`| 未找到\n| [CNN-LSTM-CTC]( https:\u002F\u002Fgithub.com\u002Foyxhust\u002FCNN-LSTM-CTC-text-recognition)  | 我实现了三种不同的文本识别模型，它们都包含 CTC 损失层，以实现对文本图像的无分割识别。   | `MXNet`| 未找到\n| [Faster_RCNN_for_DOTA]( https:\u002F\u002Fgithub.com\u002Fjessemelpolio\u002FFaster_RCNN_for_DOTA)  | 这是论文 _DOTA：航空影像中的大规模目标检测数据集_ 的官方仓库。   | `MXNet`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fjessemelpolio\u002FFaster_RCNN_for_DOTA\u002Fmaster\u002FLICENSE )\n| [RetinaNet]( https:\u002F\u002Fgithub.com\u002Funsky\u002FRetinaNet)  | 密集目标检测的焦点损失。   | `MXNet`| 未找到\n| [MobileNetV2]( https:\u002F\u002Fgithub.com\u002Fliangfu\u002Fmxnet-mobilenet-v2)  | 这是根据论文 _倒残差与线性瓶颈：用于分类、检测和分割的移动网络_ 中描述的 MobileNetV2 架构的 MXNet 实现。   | `MXNet`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fliangfu\u002Fmxnet-mobilenet-v2\u002Fmaster\u002FLICENSE )\n| [neuron-selectivity-transfer]( https:\u002F\u002Fgithub.com\u002FTuSimple\u002Fneuron-selectivity-transfer)  | 该代码是对论文 _喜欢你所喜欢的：通过神经元选择性迁移进行知识蒸馏_ 中 ImageNet 分类实验的重新实现。   | `MXNet`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002FTuSimple\u002Fneuron-selectivity-transfer\u002Fmaster\u002FLICENSE )\n| [MobileNetV2]( https:\u002F\u002Fgithub.com\u002Fchinakook\u002FMobileNetV2.mxnet)  | 这是根据论文 _倒残差与线性瓶颈：用于分类、检测和分割的移动网络_ 中描述的 MobileNetV2 架构的 Gluon 实现。   | `MXNet`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002Fchinakook\u002FMobileNetV2.mxnet\u002Fmaster\u002FLICENSE )\n| [sparse-structure-selection]( https:\u002F\u002Fgithub.com\u002FTuSimple\u002Fsparse-structure-selection)  | 该代码是对论文 _面向深度神经网络的数据驱动稀疏结构选择_ 中 ImageNet 分类实验的重新实现。   | `MXNet`| [Apache 许可证]( https:\u002F\u002Fraw.githubusercontent.com\u002FTuSimple\u002Fsparse-structure-selection\u002Fmaster\u002FLICENSE )\n| [FastPhotoStyle]( https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FFastPhotoStyle)  | 照片级真实感图像风格化的闭式解。   | `MXNet`| [知识共享署名-非商业性使用-相同方式共享 4.0 国际许可协议]( https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002FFastPhotoStyle\u002Fmaster\u002FLICENSE.md )\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#framework\">↥ 返回顶部\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n***\n\n## 贡献\n我们始终欢迎您的贡献！！\n请查看 `contributing.md`\n\n## 许可证\n\n[MIT 许可证](LICENSE)","# CV-pretrained-model 快速上手指南\n\nCV-pretrained-model 是一个汇集了多种计算机视觉（Computer Vision）预训练模型的开源资源库，涵盖目标检测、图像分割、超分辨率、图像描述生成等任务。本指南将帮助你快速了解并使用这些模型。\n\n## 环境准备\n\n在开始之前，请确保你的开发环境满足以下要求：\n\n*   **操作系统**: Linux, macOS 或 Windows\n*   **Python**: 建议版本 3.6 - 3.9 (具体版本取决于所选框架)\n*   **深度学习框架**: 根据你要使用的模型，需安装以下任一框架：\n    *   TensorFlow\n    *   Keras\n    *   PyTorch\n    *   Caffe\n    *   MXNet\n*   **硬件加速 (可选但推荐)**: NVIDIA GPU 及对应的 CUDA\u002FcuDNN 驱动，以加速模型推理和训练。\n*   **可视化工具 (可选)**: [Netron](https:\u002F\u002Fgithub.com\u002Flutzroeder\u002FNetron)，用于查看模型网络架构。\n\n### 前置依赖安装\n\n推荐使用国内镜像源加速 Python 包的安装。以下以安装 TensorFlow 和 Keras 为例（其他框架请参考官方文档）：\n\n```bash\n# 配置 pip 使用清华镜像源\npip config set global.index-url https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n\n# 安装 TensorFlow (GPU 版本示例，如需 CPU 版去掉 gpu 后缀)\npip install tensorflow-gpu\n\n# 安装 Keras (通常包含在 TensorFlow 2.x 中，也可单独安装)\npip install keras\n\n# 安装常用图像处理库\npip install opencv-python pillow matplotlib numpy\n```\n\n## 安装步骤\n\n本项目本身是一个模型列表索引，**不需要**像普通 Python 库那样通过 `pip install cv-pretrained-model` 进行安装。使用流程如下：\n\n1.  **浏览模型列表**: 访问项目的 README 或 GitHub 页面，根据任务需求（如目标检测、图像分割）和框架偏好（TensorFlow\u002FKeras\u002FPyTorch）选择合适的模型。\n2.  **克隆具体模型仓库**: 点击表格中对应模型的链接，进入其原始 GitHub 仓库。\n3.  **下载代码与权重**:\n    ```bash\n    # 示例：克隆 Mask R-CNN (Keras 版本)\n    git clone https:\u002F\u002Fgithub.com\u002Fmatterport\u002FMask_RCNN.git\n    cd Mask_RCNN\n    \n    # 安装该模型特定的依赖\n    pip install -r requirements.txt\n    ```\n    *注意：部分模型可能需要手动下载预训练权重文件（.h5, .pth, .ckpt 等），请参照各模型仓库的说明文档。*\n\n## 基本使用\n\n不同模型的具体调用方式差异较大，以下提供一个基于 **Keras** 的通用加载预训练模型（以 VGG16 为例）进行图像分类的最简示例。\n\n### 示例：使用 Keras 加载 VGG16 进行图像预测\n\n```python\nimport tensorflow as tf\nfrom tensorflow.keras.applications import VGG16\nfrom tensorflow.keras.preprocessing import image\nfrom tensorflow.keras.applications.vgg16 import preprocess_input, decode_predictions\nimport numpy as np\n\n# 1. 加载预训练模型 (权重使用 ImageNet)\n# include_top=True 表示包含全连接层，适用于分类任务\nmodel = VGG16(weights='imagenet', include_top=True)\n\n# 2. 准备输入图像\nimg_path = 'your_image.jpg'  # 替换为你的图片路径\nimg = image.load_img(img_path, target_size=(224, 224))\nx = image.img_to_array(img)\nx = np.expand_dims(x, axis=0)\nx = preprocess_input(x)\n\n# 3. 进行预测\npreds = model.predict(x)\n\n# 4. 解码预测结果\nresults = decode_predictions(preds, top=3)[0]\n\n# 5. 输出结果\nprint(\"预测结果:\")\nfor label, description, score in results:\n    print(f\"{description}: {score:.2f}\")\n```\n\n### 进阶使用提示\n\n*   **模型可视化**: 下载模型文件后，可使用 Netron 打开查看网络结构：\n    ```bash\n    netron model.h5\n    ```\n*   **迁移学习**: 大多数模型支持移除顶层（`include_top=False`），以便你在此基础上添加自定义层进行微调（Fine-tuning），适应特定领域的数据。\n*   **框架切换**: 如果首选框架不支持某模型，可查阅列表中是否提供了其他框架（如 PyTorch 或 TensorFlow）的实现版本。","某初创团队正紧急开发一款用于零售货架的自动补货检测系统，需要在两周内上线以验证商业模式。\n\n### 没有 CV-pretrained-model 时\n- **研发周期漫长**：团队需从零开始收集百万级商品图片并训练基础特征提取器，仅数据准备和模型收敛就需数月，远超项目截止日期。\n- **算力成本高昂**：从头训练深度卷积神经网络需要租用大量高性能 GPU 集群，对于资金紧张的初创公司是一笔巨大的非必要开支。\n- **技术门槛过高**：团队成员虽熟悉业务逻辑，但缺乏从头设计如 ResNet101 或 FPN 等复杂骨干网络架构的资深算法专家，导致模型精度难以达标。\n- **框架适配困难**：在尝试复现论文代码时，面临 TensorFlow、PyTorch 等不同框架的版本兼容性问题，大量时间浪费在环境调试而非业务优化上。\n\n### 使用 CV-pretrained-model 后\n- **极速启动开发**：直接调用库中基于 ImageNet 预训练的 MobileNet 或 Faster-RCNN 模型作为起点，将原本数月的冷启动时间压缩至几天内完成原型验证。\n- **显著降低成本**：利用迁移学习技术，仅需少量特定货架数据进行微调（Fine-tuning），大幅减少了对昂贵算力资源的依赖。\n- **站在巨人肩膀上**：直接复用谷歌、Facebook 等大厂开源的高精度架构（如 Mask R-CNN），确保系统在物体定位和分割任务上起步即达到行业领先水平。\n- **多框架灵活选择**：根据团队技术栈自由选择 TensorFlow 或 PyTorch 版本的预训练权重，无缝集成到现有流水线中，消除了环境适配障碍。\n\nCV-pretrained-model 通过将成熟的视觉能力转化为即插即用的模块，让开发者从重复造轮子的困境中解脱，专注于解决具体的业务难题。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fbalavenkatesh3322_CV-pretrained-model_835d61f7.png","balavenkatesh3322","Bala venkatesh","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fbalavenkatesh3322_1d460d09.jpg","Sharing trending technology concepts, ideas, and codes.",null,"India","venkateshpnk22@gmail.com","balavenkatesh22","https:\u002F\u002Fbalavenkatesh3322.github.io\u002Fbala_venkatesh_profile\u002F","https:\u002F\u002Fgithub.com\u002Fbalavenkatesh3322",1359,194,"2026-04-04T06:58:46","MIT",4,"","未说明",{"notes":91,"python":89,"dependencies":92},"该仓库是一个计算机视觉预训练模型的集合列表，而非单一的可安装工具。它列出了基于不同框架（TensorFlow, Keras, PyTorch, Caffe, MXNet）的多个独立模型项目链接。具体的运行环境需求（如操作系统、GPU、内存、Python 版本等）取决于用户选择的具体模型及其原始仓库说明。建议使用 Netron 工具可视化模型架构。",[93,94,95,96,97],"TensorFlow","Keras","PyTorch","Caffe","MXNet",[99,35,14,15,100,16],"视频","其他",[102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120],"computer-vision","tensorflow","keras","pytorch","mxnet","models","model","pretrained-models","pretrained","data-science","deep-learning","neural-network","awesome-list","object-detection","image-classification","image","video-analysis","model-selection","python3","2026-03-27T02:49:30.150509","2026-04-10T11:23:30.535956",[],[125],{"id":126,"version":127,"summary_zh":128,"released_at":129},180969,"v1.0","初始发布","2020-07-17T02:51:41"]