[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-cbsudux--awesome-human-pose-estimation":3,"tool-cbsudux--awesome-human-pose-estimation":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",138956,2,"2026-04-05T11:33:21",[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":75,"owner_avatar_url":76,"owner_bio":77,"owner_company":78,"owner_location":79,"owner_email":80,"owner_twitter":78,"owner_website":78,"owner_url":81,"languages":78,"stars":82,"forks":83,"last_commit_at":84,"license":78,"difficulty_score":85,"env_os":86,"env_gpu":87,"env_ram":87,"env_deps":88,"category_tags":91,"github_topics":92,"view_count":23,"oss_zip_url":78,"oss_zip_packed_at":78,"status":16,"created_at":101,"updated_at":102,"faqs":103,"releases":114},3390,"cbsudux\u002Fawesome-human-pose-estimation","awesome-human-pose-estimation","A collection of awesome resources in Human Pose estimation.","awesome-human-pose-estimation 是一个专注于人体姿态估计领域的优质资源合集，旨在为学习者和从业者提供一站式的技术导航。它系统性地整理了从基础概念到前沿研究的各类资料，有效解决了该领域技术更新快、论文与代码分散难寻的痛点，帮助用户快速构建知识体系并追踪最新进展。\n\n这份合集特别适合人工智能研究人员、计算机视觉开发者以及希望深入理解姿态估计技术的学生使用。无论是想入门 2D\u002F3D 姿态估算基础，还是寻找特定场景（如实时估算、多人检测）的解决方案，都能在此找到方向。其核心亮点在于结构清晰的分类索引，涵盖了经典与最新的学术论文、主流框架（PyTorch、TensorFlow 等）的代码实现、关键数据集以及深度解读博客。通过汇集带代码链接的复现资源和权威指南，awesome-human-pose-estimation 不仅降低了技术门槛，更为算法优化和项目落地提供了坚实的参考基石。","# Awesome Human Pose Estimation [![Awesome](https:\u002F\u002Fcdn.rawgit.com\u002Fsindresorhus\u002Fawesome\u002Fd7305f38d29fed78fa85652e3a63e154dd8e8829\u002Fmedia\u002Fbadge.svg)](https:\u002F\u002Fgithub.com\u002Fsindresorhus\u002Fawesome)\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fcbsudux_awesome-human-pose-estimation_readme_d05900fac496.png\" width=700>\n\u003C\u002Fp>\n\nA collection of resources on Human Pose Estimation.\n\n## Why awesome human pose estimation?\n\nThis is a collection of papers and resources I curated when learning the ropes in Human Pose estimation. I will be continuously updating this list with the latest papers and resources. If you want to learn the basics of Human Pose Estimation and understand how the field has evolved, check out these articles I published on [2D Pose Estimation](https:\u002F\u002Fblog.nanonets.com\u002Fhuman-pose-estimation-2d-guide\u002F?utm_source=github&utm_medium=social&utm_campaign=pose&utm_content=cbsudux) and [3D Pose Estimation](https:\u002F\u002Fblog.nanonets.com\u002Fhuman-pose-estimation-3d-guide\u002F)\n\n## Contributing\n\nIf you think I have missed out on something (or) have any suggestions (papers, implementations and other resources), feel free to [pull a request](https:\u002F\u002Fgithub.com\u002Fcbsudux\u002Fawesome-human-pose-estimation\u002Fpulls)\n\nFeedback and contributions are welcome!\n\n## Table of Contents\n- [Basics](#basics)\n- [Papers](#papers)\n  - [2D Pose estimation](#2d-pose-estimation)\n  - [3D Pose estimation](#3d-pose-estimation)\n  - [Person generation](#person-generation)\n  - [Real-time Pose estimation](#real-time-pose-estimation)\n- [Datasets](#datasets)\n- [Workshops](#workshops) \n- [Blog posts](#blogposts)\n- [Popular implementations](#popular-implementations)\n  - [PyTorch](#pytorch)\n  - [TensorFlow](#tensorflow)\n  - [Torch](#Torch)\n  - [Others](#others)\n\n## Basics\n- [A 2019 guide to Human Pose Estimation with Deep Learning](https:\u002F\u002Fblog.nanonets.com\u002Fhuman-pose-estimation-2d-guide\u002F?utm_source=reddit&utm_medium=social&utm_campaign=pose&utm_content=GROUP_NAME)\n\n\n## Papers\n\n### 2D Pose estimation\n- [Learning Human Pose Estimation Features with Convolutional Networks](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1312.7302.pdf) - Jain, A., Tompson, J., Andriluka, M., Taylor, G.W., & Bregler, C. (ICLR 2013) \n- [DeepPose: Human Pose Estimation via Deep Neural Networks](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1312.4659.pdf) - Toshev, A., & Szegedy, C. (CVPR 2014)\n- [Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1406.2984.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fmax-andr\u002Fjoint-cnn-mrf) - Tompson, J., Jain, A., LeCun, Y., & Bregler, C. (NIPS 2014) \n- [MoDeep: A Deep Learning Framework Using Motion Features for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1409.7963.pdf) - Jain, A., Tompson, J., LeCun, Y., & Bregler, C. (ACCV 2014)\n- [Efficient Object Localization Using Convolutional Networks](http:\u002F\u002Fciteseerx.ist.psu.edu\u002Fviewdoc\u002Fdownload?doi=10.1.1.858.5872&rep=rep1&type=pdf) - Tompson, J., Goroshin, R., Jain, A., LeCun, Y., & Bregler, C (CVPR 2015)\n- [Flowing ConvNets for Human Pose Estimation in Videos](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1506.02897.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Ftpfister\u002Fcaffe-heatmap) - Pfister, T., Charles, J., & Zisserman, A. (ICCV 2015)\n- [Convolutional Pose Machines](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1602.00134.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fshihenw\u002Fconvolutional-pose-machines-release) - Wei, S., Ramakrishna, V., Kanade, T., & Sheikh, Y. (CVPR 2016)\n- [Human Pose Estimation with Iterative Error Feedback](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1507.06550.pdf)- [[CODE]](https:\u002F\u002Fgithub.com\u002Fpulkitag\u002Fief) Carreira, J., Agrawal, P., Fragkiadaki, K., & Malik, J. (CVPR 2016) \n- [DeepCut: Joint Subset Partition and Labeling for Multi Person Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1511.06645.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Feldar\u002Fdeepcut) - Pishchulin, L., Insafutdinov, E., Tang, S., Andres, B., Andriluka, M., Gehler, P.V., & Schiele, B. (CVPR 2016)\n- [DeeperCut: A Deeper, Stronger, and Faster Multi-Person Pose Estimation Model](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1605.03170.pdf) - [[CODE1]](https:\u002F\u002Fgithub.com\u002Feldar\u002Fdeepcut-cnn)[[CODE2]](https:\u002F\u002Fgithub.com\u002Feldar\u002Fpose-tensorflow) - Insafutdinov, E., Pishchulin, L., Andres, B., Andriluka, M., & Schiele, B. (ECCV 2016)\n- [Stacked Hourglass Networks for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1603.06937.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fumich-vl\u002Fpose-hg-demo) - Newell, A., Yang, K., & Deng, J. (ECCV 2016) \n- [Multi-context Attention for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1702.07432.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fbearpaw\u002Fpose-attention) - Chu, X., Yang, W., Ouyang, W., Ma, C., Yuille, A.L., & Wang, X. (CVPR 2017)\n- [Towards Accurate Multi-person Pose Estimation in the Wild](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1701.01779.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fhackiey\u002Fkeypoints) - Papandreou, G., Zhu, T., Kanazawa, N., Toshev, A., Tompson, J., Bregler, C., & Murphy, K.P. (CVPR 2017) \n- [Realtime Multi-person 2D Pose Estimation Using Part Affinity Fields](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1611.08050.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FZheC\u002FRealtime_Multi-Person_Pose_Estimation) - Cao, Z., Simon, T., Wei, S., & Sheikh, Y. (CVPR 2017) \n- [Learning Feature Pyramids for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1708.01101.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fbearpaw\u002FPyraNet) - Yang, W., Li, S., Ouyang, W., Li, H., & Wang, X. (ICCV 2017)\n- [Human Pose Estimation Using Global and Local Normalization](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1709.07220.pdf) - Sun, K., Lan, C., Xing, J., Zeng, W., Liu, D., & Wang, J. (ICCV 2017) \n- [Adversarial PoseNet: A Structure-Aware Convolutional Network for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.00389.pdf) - Chen, Y., Shen, C., Wei, X., Liu, L., & Yang, J. (ICCV 2017)\n- [RMPE: Regional Multi-person Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1612.00137.pdf) - [[CODE1]](https:\u002F\u002Fgithub.com\u002FFang-Haoshu\u002FRMPE)[[CODE2]](https:\u002F\u002Fgithub.com\u002FMVIG-SJTU\u002FAlphaPose) - Fang, H., Xie, S., & Lu, C. (ICCV 2017)\n- [Self Adversarial Training for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1707.02439.pdf) - [[CODE1]](https:\u002F\u002Fgithub.com\u002Fdongzhuoyao\u002Fjessiechouuu-adversarial-pose)[[CODE2]](https:\u002F\u002Fgithub.com\u002Froytseng-tw\u002Fadversarial-pose-pytorch) - Chou, C., Chien, J., & Chen, H. (ArXiv 2017)\n- [Recurrent Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1605.02914.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fox-vgg\u002Fkeypoint_detection) - Belagiannis, V., & Zisserman, A. (FG 2017)\n- [Knowledge-Guided Deep Fractal Neural Networks for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.02407.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FGuanghan\u002FGNet-pose) Ning, G., Zhang, Z., & He, Z. (IEEE Transactions on Multimedia 2018)\n- [Human Pose Estimation with Parsing Induced Learner](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_cvpr_2018\u002Fpapers\u002FNie_Human_Pose_Estimation_CVPR_2018_paper.pdf)- Xuecheng Nie, Jiashi Feng, Yiming Zuo, Shuicheng Yan (CVPR 2018)\n- [LSTM Pose Machines](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.06316.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Flawy623\u002FLSTM_Pose_Machines) - Yue Luo, Jimmy Ren, Zhouxia Wang, Wenxiu Sun, Jinshan Pan, Jianbo Liu, Jiahao Pang, Liang Lin (CVPR 2018)\n- [Simple Baselines for Human Pose Estimation\nand Tracking](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FBin_Xiao_Simple_Baselines_for_ECCV_2018_paper.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002Fhuman-pose-estimation.pytorch) - Bin, Xiao, Haiping Wu, Yichen Wei (ECCV 2018)\n- [Multi-Scale Structure-Aware Network for Human Pose Estimation](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FLipeng_Ke_Multi-Scale_Structure-Aware_Network_ECCV_2018_paper.pdf) - Lipeng Ke, Ming-Ching Chang, Honggang Qi, Siwei Lyu (ECCV 2018)\n- [Deeply Learned Compositional Models for Human Pose Estimation](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FWei_Tang_Deeply_Learned_Compositional_ECCV_2018_paper.pdf) - Wei Tang, Pei Yu, Ying Wu (ECCV 2018)\n- [Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1811.12004.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation.pytorch) Osokin, D. (ArXiv 2018)\n- [Deep High-Resolution Representation Learning for Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fabs\u002F1902.09212) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fleoxiaobin\u002Fdeep-high-resolution-net.pytorch) Ke Sun, Bin Xiao, Dong Liu, Jingdong Wang (CVPR 2019)\n\n### 3D Pose estimation\n\n- [3D Human Pose Estimation from Monocular Images with Deep Convolutional Neural Network](http:\u002F\u002Fvisal.cs.cityu.edu.hk\u002Fstatic\u002Fpubs\u002Fconf\u002Faccv14-3dposecnn.pdf) - Li, S., & Chan, A.B. (ACCV 2014)\n- [Structured Prediction of 3D Human Pose with Deep Neural Networks](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1605.05180.pdf) - Tekin, B., Katircioglu, I., Salzmann, M., Lepetit, V., & Fua, P. (BMVC 2016)\n- [VNect: Real-time 3D Human Pose Estimation with a Single RGB Camera](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002Fprojects\u002FVNect\u002Fcontent\u002FVNect_SIGGRAPH2017.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Ftimctho\u002FVNect-tensorflow) - Mehta, Dushyant et al. (SIGGRAPH 2017)\n- [Recurrent 3D Pose Sequence Machines](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1707.09695.pdf) - Lin, M., Lin, L., Liang, X., Wang, K., & Cheng, H. (CVPR 2017)\n- [Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1701.00295.pdf) - Tomè, D., Russell, C., & Agapito, L. (CVPR 2017)\n- [Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1611.07828.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fgeopavlakos\u002Fc2f-vol-demo) - Pavlakos, G., Zhou, X., Derpanis, K.G., & Daniilidis, K. (CVPR 2017)\n- [Towards 3D Human Pose Estimation in the Wild: a Weakly-supervised Approach](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1704.02447.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FPytorch-pose-hg-3d) - Zhou, X., Huang, Q., Sun, X., Xue, X., & Wei, Y. (ICCV 2017)\n- [A Simple Yet Effective Baseline for 3d Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.03098.pdf) - Martinez, J., Hossain, R., Romero, J., & Little, J.J. (ICCV 2017)\n- [Compositional Human Pose Regression](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1704.00159.pdf) - Sun, X., Shang, J., Liang, S., & Wei, Y. (ICCV 2017)\n- [Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervision](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002F3dhp-dataset\u002F) - Mehta, D., Rhodin, H., Casas, D., Fua, P., Sotnychenko, O., Xu, W., & Theobalt, C. (3DV 2017)\n- [3D Human Pose Estimation in the Wild by Adversarial Learning](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1803.09722.pdf) - Yang, W., Ouyang, W., Wang, X., Ren, J.S., Li, H., & Wang, X. (2018)\n- [DRPose3D: Depth Ranking in 3D Human Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1805.08973.pdf) - Wang, M., Chen, X., Liu, W., Qian, C., Lin, L., & Ma, L. (IJCAI 2018)\n- [End-to-end Recovery of Human Shape and Pose](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.06584.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fakanazawa\u002Fhmr) - Kanazawa, A., Black, M.J., Jacobs, D.W., & Malik, J. (CVPR 2018)\n- [Learning to Estimate 3D Human Pose and Shape from a Single Color Image](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_cvpr_2018\u002Fpapers\u002FPavlakos_Learning_to_Estimate_CVPR_2018_paper.pdf) - Pavlakos, G., Zhu, L., Zhou, X., & Daniilidis, K. (CVPR 2018)\n- [Dense Human Pose Estimation In The Wild](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1802.00434.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDensepose) - Guler, R.A., Neverova, N., & Kokkinos, I. (ArXiv 2018)\n- [Neural Body Fitting: Unifying Deep Learning and Model-Based Human Pose and Shape Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1808.05942.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fmohomran\u002Fneural_body_fitting) - Omran, Mohamed and Lassner, Christoph and Pons-Moll, Gerard and Gehler, Peter V. and Schiele, Bernt (3DV 2018)\n- [Learning 3D Human Pose from Structure and Motion](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FRishabh_Dabral_Learning_3D_Human_ECCV_2018_paper.pdf) - Dabral, R., Mundhada, A., Kusupati, U., Afaque, S., Sharma, A., & Jain, A. (ECCV 2018)\n- [Integral Human Pose Regression](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1711.08229.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FJimmySuen\u002Fintegral-human-pose) - Sun, X., Xiao, B., Liang, S., & Wei, Y. (ECCV 2018)\n- [Dense Pose Transfer](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.01995.pdf) - Neverova, N., Guler, R.A., & Kokkinos, I. (ECCV 2018)\n- [Unsupervised Geometry-Aware Representation for 3D Human Pose Estimation](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FHelge_Rhodin_Unsupervised_Geometry-Aware_Representation_ECCV_2018_paper.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fhrhodin\u002FUnsupervisedGeometryAwareRepresentationLearning) - Rhodin, H., Salzmann, M., & Fua, P. (ECCV 2018)\n- [BodyNet: Volumetric Inference of 3D Human Body Shapes](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1804.04875v3.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fgulvarol\u002Fbodynet) - Varol, G., Ceylan, D., Russell, B., Yang, J., Yumer, E., Laptev, I., & Schmid, C. (ECCV 2018)\n- [3D human pose estimation in video with temporal convolutions and\nsemi-supervised training](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1811.11742.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FVideoPose3D) - Pavllo, D., Feichtenhofer, C., Grangier, D., & Auli, M (ArXiv 2018)\n- [Single-Shot Multi-Person 3D Pose Estimation From Monocular RGB](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.03453.pdf) - [[CODE1]](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002Fprojects\u002FSingleShotMultiPerson\u002F)[[CODE2]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation-3d-demo.pytorch) - Mehta, D.,  Sotnychenko, O., Mueller, F., Xu, W., Sridhar, S., Pons-Moll, G., Theobalt, C. (3DV 2018)\n\n\n### Person generation\n\n- [Pose Guided Person Image Generation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.09368.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Fcharliememory\u002FPose-Guided-Person-Image-Generation) - Ma, L., Jia, X., Sun, Q., Schiele, B., Tuytelaars, T., & Gool, L.V. (NIPS 2017)\n- [A Generative Model of People in Clothing](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.04098.pdf) - Lassner, C., Pons-Moll, G., & Gehler, P.V. (ICCV 2017)\n- [Deformable GANs for Pose-based Human Image Generation](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_cvpr_2018\u002Fpapers\u002FSiarohin_Deformable_GANs_for_CVPR_2018_paper.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FAliaksandrSiarohin\u002Fpose-gan) - Siarohin, A., Sangineto, E., Lathuilière, S., & Sebe, N. (CVPR 2018)\n- [Dense Pose Transfer](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.01995.pdf) - Neverova, N., Guler, R.A., & Kokkinos, I. (ECCV 2018)\n\n### Real-time pose estimation\n\n\n- [Realtime Multi-person 2D Pose Estimation Using Part Affinity Fields](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1611.08050.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FZheC\u002FRealtime_Multi-Person_Pose_Estimation) - Cao, Z., Simon, T., Wei, S., & Sheikh, Y. (CVPR 2017) \n- [VNect: Real-time 3D Human Pose Estimation with a Single RGB Camera](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002Fprojects\u002FVNect\u002Fcontent\u002FVNect_SIGGRAPH2017.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Ftimctho\u002FVNect-tensorflow) - Mehta, Dushyant et al. (SIGGRAPH 2017)\n- [RMPE: Regional Multi-person Pose Estimation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1612.00137.pdf) - [[CODE1]](https:\u002F\u002Fgithub.com\u002FFang-Haoshu\u002FRMPE)[[CODE2]](https:\u002F\u002Fgithub.com\u002FMVIG-SJTU\u002FAlphaPose) - Fang, H., Xie, S., & Lu, C. (ICCV 2017)\n- [Dense Human Pose Estimation In The Wild](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1802.00434.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDensepose) - Guler, R.A., Neverova, N., & Kokkinos, I. (ArXiv 2018)\n- [Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1811.12004.pdf) - [[CODE]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation.pytorch) Osokin, D. (ArXiv 2018)\n  - Extension to 3D pose estimation (based on [Single-Shot Multi-Person 3D Pose Estimation From Monocular RGB](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.03453.pdf) - Mehta, D., et al.) - [[CODE]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation-3d-demo.pytorch)\n\n\n\n\n## Datasets\n### 2D\n- [MPII Human Pose Dataset](http:\u002F\u002Fhuman-pose.mpi-inf.mpg.de\u002F)\n- [LSP](http:\u002F\u002Fsam.johnson.io\u002Fresearch\u002Flsp.html)\n- [FLIC](https:\u002F\u002Fbensapp.github.io\u002Fflic-dataset.html)\n- [FLIC-plus](https:\u002F\u002Fcims.nyu.edu\u002F~tompson\u002Fflic_plus.htm)\n\n### 3D\n- [Human3.6M](http:\u002F\u002Fvision.imar.ro\u002Fhuman3.6m\u002Fdescription.php)\n- [HumanEva](http:\u002F\u002Fhumaneva.is.tue.mpg.de\u002F)\n- [MPI-INF-3DHP](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002F3dhp-dataset\u002F)\n- [Unite The People](http:\u002F\u002Ffiles.is.tuebingen.mpg.de\u002Fclassner\u002Fup\u002F)\n\n\n## Workshops\n- [POSETRACK-ECCV](https:\u002F\u002Fposetrack.net\u002Fworkshops\u002Feccv2018\u002F)\n- [3D HUMANS-CVPR 2018](https:\u002F\u002Fproject.inria.fr\u002Fhumans2018\u002F)\n\n\n## Blog posts\n- [Real-time Human Pose Estimation in the Browser with TensorFlow.js](https:\u002F\u002Fmedium.com\u002Ftensorflow\u002Freal-time-human-pose-estimation-in-the-browser-with-tensorflow-js-7dd0bc881cd5)\n- [Deep learning for human pose estimation](https:\u002F\u002Fwww.slideshare.net\u002Fplutoyang\u002Fmmlab-seminar-2016-deep-learning-for-human-pose-estimation)\n- [Deep Learning based Human Pose Estimation using OpenCV ( C++ \u002F Python )](https:\u002F\u002Fwww.learnopencv.com\u002Fdeep-learning-based-human-pose-estimation-using-opencv-cpp-python\u002F)\n\n## Popular implementations\n\n\n### PyTorch\n- [pytorch-pose-hg-3d](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FPytorch-pose-hg-3d)\n- [3d_pose_baseline_pytorch](https:\u002F\u002Fgithub.com\u002Fweigq\u002F3d_pose_baseline_pytorch)\n- [pytorch_Realtime_Multi-Person_Pose_Estimation](https:\u002F\u002Fgithub.com\u002Ftensorboy\u002Fpytorch_Realtime_Multi-Person_Pose_Estimation)\n- [AlphaPose](https:\u002F\u002Fgithub.com\u002FMVIG-SJTU\u002FAlphaPose\u002Ftree\u002Fpytorch)\n- [pytorch-pose](https:\u002F\u002Fgithub.com\u002Fbearpaw\u002Fpytorch-pose)\n- [human-pose-estimation.pytorch](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002Fhuman-pose-estimation.pytorch)\n- [deep-high-resolution-net.pytorch](https:\u002F\u002Fgithub.com\u002Fleoxiaobin\u002Fdeep-high-resolution-net.pytorch)\n\n### TensorFlow\n\n- [tf-pose-estimation](https:\u002F\u002Fgithub.com\u002Fildoonet\u002Ftf-pose-estimation)\n- [pose-tensorflow](https:\u002F\u002Fgithub.com\u002Feldar\u002Fpose-tensorflow)\n\n### Torch\n\n- [pose-hg-train](https:\u002F\u002Fgithub.com\u002Fumich-vl\u002Fpose-hg-train)\n- [pose-hg-demo](https:\u002F\u002Fgithub.com\u002Fumich-vl\u002Fpose-hg-demo)\n\n### Others\n\n- [openpose](https:\u002F\u002Fgithub.com\u002FCMU-Perceptual-Computing-Lab\u002Fopenpose)\n- [DensePose](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDensePose)\n\n## Todo\n\n- [x] Add basics\n- [ ] Add papers on Person Re-Identification\n- [ ] Add papers on Multi Person Pose estimation\n- [ ] Add a SOTA ranking\n\n## License\n\n\u003Ca rel=\"license\" href=\"http:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby\u002F4.0\u002F\">\u003Cimg alt=\"Creative Commons License\" style=\"border-width:0\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fcbsudux_awesome-human-pose-estimation_readme_4650c94e56fa.png\" \u002F>\u003C\u002Fa>\u003Cbr \u002F>This work is licensed under a \u003Ca rel=\"license\" href=\"http:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby\u002F4.0\u002F\">Creative Commons Attribution 4.0 International License\u003C\u002Fa>.\n\n\n\n\n\n\n\n\n\n\n","# 优秀的人体姿态估计 [![Awesome](https:\u002F\u002Fcdn.rawgit.com\u002Fsindresorhus\u002Fawesome\u002Fd7305f38d29fed78fa85652e3a63e154dd8e8829\u002Fmedia\u002Fbadge.svg)](https:\u002F\u002Fgithub.com\u002Fsindresorhus\u002Fawesome)\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fcbsudux_awesome-human-pose-estimation_readme_d05900fac496.png\" width=700>\n\u003C\u002Fp>\n\n人体姿态估计相关资源的集合。\n\n## 为什么是优秀的人体姿态估计？\n\n这是我学习人体姿态估计时整理的一系列论文和资源。我会持续更新这份列表，加入最新的论文和资源。如果你想了解人体姿态估计的基础知识，并理解该领域的演进历程，请查看我在[2D姿态估计](https:\u002F\u002Fblog.nanonets.com\u002Fhuman-pose-estimation-2d-guide\u002F?utm_source=github&utm_medium=social&utm_campaign=pose&utm_content=cbsudux)和[3D姿态估计](https:\u002F\u002Fblog.nanonets.com\u002Fhuman-pose-estimation-3d-guide\u002F)上发表的文章。\n\n## 贡献\n\n如果你认为我遗漏了某些内容，或者有任何建议（例如论文、实现和其他资源），欢迎随时[提交拉取请求](https:\u002F\u002Fgithub.com\u002Fcbsudux\u002Fawesome-human-pose-estimation\u002Fpulls)。\n\n我们非常欢迎反馈和贡献！\n\n## 目录\n- [基础知识](#basics)\n- [论文](#papers)\n  - [2D姿态估计](#2d-pose-estimation)\n  - [3D姿态估计](#3d-pose-estimation)\n  - [人物生成](#person-generation)\n  - [实时姿态估计](#real-time-pose-estimation)\n- [数据集](#datasets)\n- [研讨会](#workshops) \n- [博客文章](#blogposts)\n- [流行实现](#popular-implementations)\n  - [PyTorch](#pytorch)\n  - [TensorFlow](#tensorflow)\n  - [Torch](#Torch)\n  - [其他](#others)\n\n## 基础知识\n- [2019年深度学习在人体姿态估计中的指南](https:\u002F\u002Fblog.nanonets.com\u002Fhuman-pose-estimation-2d-guide\u002F?utm_source=reddit&utm_medium=social&utm_campaign=pose&utm_content=GROUP_NAME)\n\n\n## 论文\n\n### 2D姿态估计\n- [利用卷积网络学习人体姿态估计特征](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1312.7302.pdf) - Jain, A., Tompson, J., Andriluka, M., Taylor, G.W., & Bregler, C. (ICLR 2013)\n- [DeepPose：基于深度神经网络的人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1312.4659.pdf) - Toshev, A., & Szegedy, C. (CVPR 2014)\n- [卷积网络与图模型联合训练用于人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1406.2984.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fmax-andr\u002Fjoint-cnn-mrf) - Tompson, J., Jain, A., LeCun, Y., & Bregler, C. (NIPS 2014)\n- [MoDeep：一种使用运动特征进行人体姿态估计的深度学习框架](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1409.7963.pdf) - Jain, A., Tompson, J., LeCun, Y., & Bregler, C. (ACCV 2014)\n- [利用卷积网络进行高效的目标定位](http:\u002F\u002Fciteseerx.ist.psu.edu\u002Fviewdoc\u002Fdownload?doi=10.1.1.858.5872&rep=rep1&type=pdf) - Tompson, J., Goroshin, R., Jain, A., LeCun, Y., & Bregler, C (CVPR 2015)\n- [用于视频中人体姿态估计的流式卷积网络](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1506.02897.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Ftpfister\u002Fcaffe-heatmap) - Pfister, T., Charles, J., & Zisserman, A. (ICCV 2015)\n- [卷积姿态机器](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1602.00134.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fshihenw\u002Fconvolutional-pose-machines-release) - Wei, S., Ramakrishna, V., Kanade, T., & Sheikh, Y. (CVPR 2016)\n- [基于迭代误差反馈的人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1507.06550.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fpulkitag\u002Fief) Carreira, J., Agrawal, P., Fragkiadaki, K., & Malik, J. (CVPR 2016)\n- [DeepCut：多人姿态估计中的联合子集划分与标注](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1511.06645.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Feldar\u002Fdeepcut) - Pishchulin, L., Insafutdinov, E., Tang, S., Andres, B., Andriluka, M., Gehler, P.V., & Schiele, B. (CVPR 2016)\n- [DeeperCut：更深入、更强、更快的多人姿态估计模型](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1605.03170.pdf) - [[代码1]](https:\u002F\u002Fgithub.com\u002Feldar\u002Fdeepcut-cnn)[[代码2]](https:\u002F\u002Fgithub.com\u002Feldar\u002Fpose-tensorflow) - Insafutdinov, E., Pishchulin, L., Andres, B., Andriluka, M., & Schiele, B. (ECCV 2016)\n- [用于人体姿态估计的堆叠沙漏网络](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1603.06937.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fumich-vl\u002Fpose-hg-demo) - Newell, A., Yang, K., & Deng, J. (ECCV 2016)\n- [用于人体姿态估计的多上下文注意力机制](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1702.07432.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fbearpaw\u002Fpose-attention) - Chu, X., Yang, W., Ouyang, W., Ma, C., Yuille, A.L., & Wang, X. (CVPR 2017)\n- [迈向野外环境中精确的多人姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1701.01779.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fhackiey\u002Fkeypoints) - Papandreou, G., Zhu, T., Kanazawa, N., Toshev, A., Tompson, J., Bregler, C., & Murphy, K.P. (CVPR 2017)\n- [基于部分亲和力场的实时多人2D姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1611.08050.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FZheC\u002FRealtime_Multi-Person_Pose_Estimation) - Cao, Z., Simon, T., Wei, S., & Sheikh, Y. (CVPR 2017)\n- [用于人体姿态估计的特征金字塔学习](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1708.01101.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fbearpaw\u002FPyraNet) - Yang, W., Li, S., Ouyang, W., Li, H., & Wang, X. (ICCV 2017)\n- [采用全局与局部归一化的人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1709.07220.pdf) - Sun, K., Lan, C., Xing, J., Zeng, W., Liu, D., & Wang, J. (ICCV 2017)\n- [对抗姿态网络：一种结构感知的卷积网络用于人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.00389.pdf) - Chen, Y., Shen, C., Wei, X., Liu, L., & Yang, J. (ICCV 2017)\n- [RMPE：区域多人姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1612.00137.pdf) - [[代码1]](https:\u002F\u002Fgithub.com\u002FFang-Haoshu\u002FRMPE)[[代码2]](https:\u002F\u002Fgithub.com\u002FMVIG-SJTU\u002FAlphaPose) - Fang, H., Xie, S., & Lu, C. (ICCV 2017)\n- [用于人体姿态估计的自对抗训练](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1707.02439.pdf) - [[代码1]](https:\u002F\u002Fgithub.com\u002Fdongzhuoyao\u002Fjessiechouuu-adversarial-pose)[[代码2]](https:\u002F\u002Fgithub.com\u002Froytseng-tw\u002Fadversarial-pose-pytorch) - Chou, C., Chien, J., & Chen, H. (ArXiv 2017)\n- [循环人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1605.02914.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fox-vgg\u002Fkeypoint_detection) - Belagiannis, V., & Zisserman, A. (FG 2017)\n- [知识引导的深度分形神经网络用于人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.02407.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FGuanghan\u002FGNet-pose) Ning, G., Zhang, Z., & He, Z. (IEEE Transactions on Multimedia 2018)\n- [基于解析诱导学习器的人体姿态估计](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_cvpr_2018\u002Fpapers\u002FNie_Human_Pose_Estimation_CVPR_2018_paper.pdf) - Xuecheng Nie, Jiashi Feng, Yiming Zuo, Shuicheng Yan (CVPR 2018)\n- [LSTM姿态机器](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.06316.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Flawy623\u002FLSTM_Pose_Machines) - Yue Luo, Jimmy Ren, Zhouxia Wang, Wenxiu Sun, Jinshan Pan, Jianbo Liu, Jiahao Pang, Liang Lin (CVPR 2018)\n- [用于人体姿态估计与跟踪的简单基线](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FBin_Xiao_Simple_Baselines_for_ECCV_2018_paper.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002Fhuman-pose-estimation.pytorch) - Bin, Xiao, Haiping Wu, Yichen Wei (ECCV 2018)\n- [用于人体姿态估计的多尺度结构感知网络](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FLipeng_Ke_Multi-Scale_Structure-Aware_Network_ECCV_2018_paper.pdf) - Lipeng Ke, Ming-Ching Chang, Honggang Qi, Siwei Lyu (ECCV 2018)\n- [用于人体姿态估计的深度学习组合模型](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FWei_Tang_Deeply_Learned_Compositional_ECCV_2018_paper.pdf) - Wei Tang, Pei Yu, Ying Wu (ECCV 2018)\n- [在CPU上实现的实时2D多人姿态估计：轻量级OpenPose](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1811.12004.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation.pytorch) Osokin, D. (ArXiv 2018)\n- [用于人体姿态估计的深度高分辨率表征学习](https:\u002F\u002Farxiv.org\u002Fabs\u002F1902.09212) - [[代码]](https:\u002F\u002Fgithub.com\u002Fleoxiaobin\u002Fdeep-high-resolution-net.pytorch) Ke Sun, Bin Xiao, Dong Liu, Jingdong Wang (CVPR 2019)\n\n### 3D 姿态估计\n\n- [基于深度卷积神经网络的单目图像3D人体姿态估计](http:\u002F\u002Fvisal.cs.cityu.edu.hk\u002Fstatic\u002Fpubs\u002Fconf\u002Faccv14-3dposecnn.pdf) - 李思远、陈安邦（ACCV 2014）\n- [利用深度神经网络进行3D人体姿态的结构化预测](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1605.05180.pdf) - 特金·巴哈迪尔、卡蒂尔焦卢·伊尔汗、萨尔茨曼·马库斯、勒佩蒂·文森特、富阿·皮埃尔（BMVC 2016）\n- [VNect：基于单个RGB摄像头的实时3D人体姿态估计](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002Fprojects\u002FVNect\u002Fcontent\u002FVNect_SIGGRAPH2017.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Ftimctho\u002FVNect-tensorflow) - 迪舒扬特·梅塔等（SIGGRAPH 2017）\n- [循环3D姿态序列模型](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1707.09695.pdf) - 林茂、林立、梁晓轩、王凯、程浩（CVPR 2017）\n- [从深层特征中提升：基于单张图像的卷积3D姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1701.00295.pdf) - 托梅·达尼埃莱、拉塞尔·克里斯托弗、阿加皮托·路易斯（CVPR 2017）\n- [单目3D人体姿态的粗到精体积预测](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1611.07828.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fgeopavlakos\u002Fc2f-vol-demo) - 帕夫拉科斯·格奥尔吉奥斯、周晓东、德尔帕尼斯·康斯坦丁、达尼利迪斯·基里亚科斯（CVPR 2017）\n- [迈向野外环境下的3D人体姿态估计：一种弱监督方法](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1704.02447.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FPytorch-pose-hg-3d) - 周晓东、黄琦、孙旭、薛雪、魏勇（ICCV 2017）\n- [一种简单而有效的3D人体姿态估计基线](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.03098.pdf) - 马丁内斯·豪尔赫、侯赛因·拉希德、罗梅罗·胡安、利特尔·约翰（ICCV 2017）\n- [组合式人体姿态回归](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1704.00159.pdf) - 孙旭、尚杰、梁松、魏勇（ICCV 2017）\n- [利用改进的CNN监督技术进行野外环境下的单目3D人体姿态估计](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002F3dhp-dataset\u002F) - 梅塔·迪舒扬特、罗丁·赫尔格、卡萨斯·大卫、富阿·皮埃尔、索特尼琴科·奥列克桑德、徐伟、特奥巴尔特·克里斯蒂安（3DV 2017）\n- [通过对抗学习实现野外环境下的3D人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1803.09722.pdf) - 杨伟、欧阳旺、王欣、任继生、李华、王翔（2018年）\n- [DRPose3D：3D人体姿态估计中的深度排序](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1805.08973.pdf) - 王明、陈曦、刘伟、钱灿、林立、马丽（IJCAI 2018）\n- [端到端的人体形状与姿态恢复](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.06584.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fakanazawa\u002Fhmr) - 金泽步、布莱克·迈克尔、雅各布斯·戴维、马利克·贾米勒（CVPR 2018）\n- [从单张彩色图像中学习估计3D人体姿态与形状](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_cvpr_2018\u002Fpapers\u002FPavlakos_Learning_to_Estimate_CVPR_2018_paper.pdf) - 帕夫拉科斯·格奥尔吉奥斯、朱莉娅·祖、周晓东、达尼利迪斯·基里亚科斯（CVPR 2018）\n- [野外环境下的密集人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1802.00434.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDensepose) - 古勒·雷扎、涅韦罗娃·纳塔莉娅、科基诺斯·伊万尼斯（ArXiv 2018）\n- [神经人体拟合：统一深度学习与基于模型的人体姿态和形状估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1808.05942.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fmohomran\u002Fneural_body_fitting) - 奥姆兰·穆罕默德、拉斯纳·克里斯托夫、蓬斯-莫尔·杰拉尔德、盖勒·彼得、席勒·伯恩特（3DV 2018）\n- [从结构与运动中学习3D人体姿态](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FRishabh_Dabral_Learning_3D_Human_ECCV_2018_paper.pdf) - 达布拉尔·里沙布、蒙达达·阿努普、库苏帕蒂·乌玛、阿法克·萨米尔、夏尔马·阿努什、贾因·阿努普（ECCV 2018）\n- [积分式人体姿态回归](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1711.08229.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FJimmySuen\u002Fintegral-human-pose) - 孙旭、肖斌、梁松、魏勇（ECCV 2018）\n- [密集姿态迁移](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.01995.pdf) - 涅韦罗娃·纳塔莉娅、古勒·雷扎、科基诺斯·伊万尼斯（ECCV 2018）\n- [用于3D人体姿态估计的无监督几何感知表示](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_ECCV_2018\u002Fpapers\u002FHelge_Rhodin_Unsupervised_Geometry-Aware_Representation_ECCV_2018_paper.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fhrhodin\u002FUnsupervisedGeometryAwareRepresentationLearning) - 罗丁·赫尔格、萨尔茨曼·马库斯、富阿·皮埃尔（ECCV 2018）\n- [BodyNet：3D人体形态的体积推理](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1804.04875v3.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fgulvarol\u002Fbodynet) - 瓦罗尔·古尔坎、塞伊兰·代尼兹、拉塞尔·布赖恩、杨洁、尤默尔·艾伦、拉普捷夫·伊戈尔、施密德·克里斯蒂安（ECCV 2018）\n- [利用时间卷积和半监督训练在视频中进行3D人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1811.11742.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FVideoPose3D) - 帕夫洛·丹尼尔、费希滕霍费尔·克里斯蒂安、格朗吉耶·多米尼克、奥利·马泰乌斯（ArXiv 2018）\n- [基于单目RGB图像的单次多人群3D姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.03453.pdf) - [[代码1]](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002Fprojects\u002FSingleShotMultiPerson\u002F)[[代码2]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation-3d-demo.pytorch) - 梅塔·迪舒扬特、索特尼琴科·奥列克桑德、穆勒·弗里德里希、徐伟、斯里达尔·萨蒂什、蓬斯-莫尔·杰拉尔德、特奥巴尔特·克里斯蒂安（3DV 2018）\n\n\n### 人物生成\n\n- [姿态引导的人物图像生成](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.09368.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Fcharliememory\u002FPose-Guided-Person-Image-Generation) - 马丽、贾霞、孙青、席勒·伯恩特、图伊特拉尔斯·蒂恩、古尔·莱昂·范（NIPS 2017）\n- [服装中人物的生成模型](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.04098.pdf) - 拉斯纳·克里斯托夫、蓬斯-莫尔·杰拉尔德、盖勒·彼得·范（ICCV 2017）\n- [基于姿态的人像生成可变形GAN](http:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_cvpr_2018\u002Fpapers\u002FSiarohin_Deformable_GANs_for_CVPR_2018_paper.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FAliaksandrSiarohin\u002Fpose-gan) - 西亚罗欣·阿利亚克桑德尔、桑吉内托·埃马努埃莱、拉图伊利耶·塞巴斯蒂安、塞贝·尼尔（CVPR 2018）\n- [密集姿态迁移](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.01995.pdf) - 涅韦罗娃·纳塔莉娅、古勒·雷扎、科基诺斯·伊万尼斯（ECCV 2018）\n\n### 实时姿态估计\n\n\n- [基于部件亲和场的实时多人2D姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1611.08050.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FZheC\u002FRealtime_Multi-Person_Pose_Estimation) - Cao, Z., Simon, T., Wei, S., & Sheikh, Y. (CVPR 2017) \n- [VNect：使用单个RGB摄像头的实时3D人体姿态估计](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002Fprojects\u002FVNect\u002Fcontent\u002FVNect_SIGGRAPH2017.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Ftimctho\u002FVNect-tensorflow) - Mehta, Dushyant 等 (SIGGRAPH 2017)\n- [RMPE：区域多人姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1612.00137.pdf) - [[代码1]](https:\u002F\u002Fgithub.com\u002FFang-Haoshu\u002FRMPE)[[代码2]](https:\u002F\u002Fgithub.com\u002FMVIG-SJTU\u002FAlphaPose) - Fang, H., Xie, S., & Lu, C. (ICCV 2017)\n- [野外密集人体姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1802.00434.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDensepose) - Guler, R.A., Neverova, N., & Kokkinos, I. (ArXiv 2018)\n- [在CPU上进行实时2D多人姿态估计：轻量级OpenPose](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1811.12004.pdf) - [[代码]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation.pytorch) Osokin, D. (ArXiv 2018)\n  - 扩展到3D姿态估计（基于[单目RGB图像的单次多人3D姿态估计](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.03453.pdf) - Mehta, D., et al.) - [[代码]](https:\u002F\u002Fgithub.com\u002FDaniil-Osokin\u002Flightweight-human-pose-estimation-3d-demo.pytorch)\n\n\n\n\n## 数据集\n### 2D\n- [MPII人体姿态数据集](http:\u002F\u002Fhuman-pose.mpi-inf.mpg.de\u002F)\n- [LSP](http:\u002F\u002Fsam.johnson.io\u002Fresearch\u002Flsp.html)\n- [FLIC](https:\u002F\u002Fbensapp.github.io\u002Fflic-dataset.html)\n- [FLIC-plus](https:\u002F\u002Fcims.nyu.edu\u002F~tompson\u002Fflic_plus.htm)\n\n### 3D\n- [Human3.6M](http:\u002F\u002Fvision.imar.ro\u002Fhuman3.6m\u002Fdescription.php)\n- [HumanEva](http:\u002F\u002Fhumaneva.is.tue.mpg.de\u002F)\n- [MPI-INF-3DHP](http:\u002F\u002Fgvv.mpi-inf.mpg.de\u002F3dhp-dataset\u002F)\n- [Unite The People](http:\u002F\u002Ffiles.is.tuebingen.mpg.de\u002Fclassner\u002Fup\u002F)\n\n\n## 研讨会\n- [POSETRACK-ECCV](https:\u002F\u002Fposetrack.net\u002Fworkshops\u002Feccv2018\u002F)\n- [3D HUMANS-CVPR 2018](https:\u002F\u002Fproject.inria.fr\u002Fhumans2018\u002F)\n\n\n## 博客文章\n- [使用TensorFlow.js在浏览器中进行实时人体姿态估计](https:\u002F\u002Fmedium.com\u002Ftensorflow\u002Freal-time-human-pose-estimation-in-the-browser-with-tensorflow-js-7dd0bc881cd5)\n- [用于人体姿态估计的深度学习](https:\u002F\u002Fwww.slideshare.net\u002Fplutoyang\u002Fmmlab-seminar-2016-deep-learning-for-human-pose-estimation)\n- [基于深度学习的人体姿态估计：使用OpenCV（C++\u002FPython）](https:\u002F\u002Fwww.learnopencv.com\u002Fdeep-learning-based-human-pose-estimation-using-opencv-cpp-python\u002F)\n\n## 流行的实现\n\n\n### PyTorch\n- [pytorch-pose-hg-3d](https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002FPytorch-pose-hg-3d)\n- [3d_pose_baseline_pytorch](https:\u002F\u002Fgithub.com\u002Fweigq\u002F3d_pose_baseline_pytorch)\n- [pytorch_Realtime_Multi-Person_Pose_Estimation](https:\u002F\u002Fgithub.com\u002Ftensorboy\u002Fpytorch_Realtime_Multi-Person_Pose_Estimation)\n- [AlphaPose](https:\u002F\u002Fgithub.com\u002FMVIG-SJTU\u002FAlphaPose\u002Ftree\u002Fpytorch)\n- [pytorch-pose](https:\u002F\u002Fgithub.com\u002Fbearpaw\u002Fpytorch-pose)\n- [human-pose-estimation.pytorch](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002Fhuman-pose-estimation.pytorch)\n- [deep-high-resolution-net.pytorch](https:\u002F\u002Fgithub.com\u002Fleoxiaobin\u002Fdeep-high-resolution-net.pytorch)\n\n### TensorFlow\n\n- [tf-pose-estimation](https:\u002F\u002Fgithub.com\u002Fildoonet\u002Ftf-pose-estimation)\n- [pose-tensorflow](https:\u002F\u002Fgithub.com\u002Feldar\u002Fpose-tensorflow)\n\n### Torch\n\n- [pose-hg-train](https:\u002F\u002Fgithub.com\u002Fumich-vl\u002Fpose-hg-train)\n- [pose-hg-demo](https:\u002F\u002Fgithub.com\u002Fumich-vl\u002Fpose-hg-demo)\n\n### 其他\n\n- [openpose](https:\u002F\u002Fgithub.com\u002FCMU-Perceptual-Computing-Lab\u002Fopenpose)\n- [DensePose](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FDensePose)\n\n## 待办事项\n\n- [x] 添加基础知识\n- [ ] 添加关于行人再识别的论文\n- [ ] 添加关于多人姿态估计的论文\n- [ ] 添加SOTA排名\n\n## 许可证\n\n\u003Ca rel=\"license\" href=\"http:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby\u002F4.0\u002F\">\u003Cimg alt=\"知识共享许可\" style=\"border-width:0\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fcbsudux_awesome-human-pose-estimation_readme_4650c94e56fa.png\" \u002F>\u003C\u002Fa>\u003Cbr \u002F>本作品采用\u003Ca rel=\"license\" href=\"http:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby\u002F4.0\u002F\">知识共享署名4.0国际许可协议\u003C\u002Fa>授权。","# Awesome Human Pose Estimation 快速上手指南\n\n`awesome-human-pose-estimation` 并非一个单一的可执行软件或 Python 包，而是一个 curated（精选）的资源列表，汇集了人体姿态估计领域的经典论文、数据集、博客教程以及基于不同框架（PyTorch, TensorFlow 等）的代码实现。\n\n本指南将指导你如何利用该列表中的资源，快速搭建并运行一个主流的人体姿态估计模型（以列表中推荐的 **Simple Baselines (PyTorch)** 为例）。\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 (推荐 Ubuntu 18.04\u002F20.04) 或 macOS。Windows 用户建议使用 WSL2。\n*   **Python**: 3.6 或更高版本。\n*   **GPU**: 推荐使用 NVIDIA GPU 以加速推理和训练（需安装对应的 CUDA 驱动）。\n*   **前置依赖**:\n    *   Git\n    *   pip 或 conda (推荐使用 conda 管理环境)\n\n## 安装步骤\n\n由于该仓库是资源索引，我们需要从中选择一个具体的实现进行安装。这里选择列表中 **Popular implementations -> PyTorch** 部分提到的经典项目：**Simple Baselines for Human Pose Estimation**。\n\n### 1. 克隆代码仓库\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FMicrosoft\u002Fhuman-pose-estimation.pytorch.git\ncd human-pose-estimation.pytorch\n```\n\n### 2. 创建虚拟环境\n建议使用 Conda 创建隔离环境，避免依赖冲突。\n```bash\nconda create -n pose_estimation python=3.7\nconda activate pose_estimation\n```\n\n### 3. 安装依赖库\n该项目依赖 PyTorch 和其他计算机视觉库。国内用户推荐使用清华源或阿里源加速安装。\n\n**安装 PyTorch (以 CUDA 11.3 为例，请根据实际显卡调整):**\n```bash\npip install torch torchvision torchaudio --index-url https:\u002F\u002Fdownload.pytorch.org\u002Fwhl\u002Fcu113\n# 或者使用清华镜像源\n# pip install torch torchvision torchaudio --index-url https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n```\n\n**安装其他依赖:**\n```bash\npip install -r requirements.txt\n```\n*注：如果 `requirements.txt` 缺失或报错，通常需手动安装 `opencv-python`, `yacs`, `tqdm`, `numpy` 等核心库。*\n\n### 4. 下载预训练模型\n从原项目 Release 页面或列表中提供的链接下载预训练权重文件（例如 `pose_coco.pth.tar`），并将其放置在 `models\u002Fcheckpoints\u002F` 目录下（具体路径请参考项目内的 `README` 说明）。\n\n## 基本使用\n\n以下是最简单的单张图片推理示例，用于检测图片中的人体关键点。\n\n### 1. 准备测试图片\n将一张包含人物的图片（如 `test.jpg`）放入项目根目录或指定文件夹。\n\n### 2. 运行推理脚本\n使用提供的测试脚本加载模型并进行预测。以下命令假设你已配置好模型路径和数据集格式（COCO 格式）：\n\n```bash\npython tools\u002Ftest.py \\\n    --cfg experiments\u002Fcoco\u002Fresnet\u002Fpose_resnet_50_256x192.yaml \\\n    TEST.MODEL_FILE models\u002Fcheckpoints\u002Fpose_coco.pth.tar \\\n    TEST.USE_GT_BBOX False\n```\n\n*   `--cfg`: 指定配置文件，定义了网络结构（如 ResNet-50）和输入尺寸。\n*   `TEST.MODEL_FILE`: 指定下载的预训练权重路径。\n*   `TEST.USE_GT_BBOX False`: 表示不使用真实标注框，让模型自动检测人物区域。\n\n### 3. 查看结果\n运行结束后，生成的带有骨架关键点标注的图片通常保存在 `output` 或 `vis` 文件夹中。你可以使用图片查看器打开结果图，验证姿态估计效果。\n\n---\n*提示：若要尝试 3D 姿态估计或其他架构（如 HRNet, OpenPose），请回到 `awesome-human-pose-estimation` 列表的对应章节，查找相关论文的官方代码链接并重复上述“克隆 - 安装 - 运行”流程。*","某智慧体育科技公司的算法团队正致力于开发一款基于手机摄像头的 AI 健身动作纠正应用，需要快速构建高精度的人体姿态识别模型。\n\n### 没有 awesome-human-pose-estimation 时\n- **文献检索如大海捞针**：团队成员需在 arXiv、Google Scholar 等多个平台分散搜索，难以系统掌握从 DeepPose 到 CPM 等经典算法的演进脉络。\n- **复现成本极高**：寻找论文对应的开源代码耗时费力，常遇到代码缺失、框架版本不兼容或缺乏预训练模型的问题，导致基础验证周期长达数周。\n- **技术选型盲目**：缺乏对 2D 与 3D 姿态估计、实时检测等不同细分领域主流方案的横向对比，容易选错不适合移动端部署的技术路线。\n- **数据资源匮乏**：难以快速定位高质量的标准数据集（如 MPII、COCO），导致模型训练初期因数据清洗和标注问题陷入停滞。\n\n### 使用 awesome-human-pose-estimation 后\n- **知识体系一键构建**：直接利用整理好的分类目录，团队在半天内便梳理清楚技术发展史，并锁定了适合移动端的轻量级实时估算方案。\n- **工程落地加速**：通过列表中提供的 PyTorch 和 TensorFlow 热门实现链接及对应代码库，直接复用成熟模块，将原型开发时间从数周缩短至 3 天。\n- **决策依据充分**：参考收录的权威论文与博客解读，团队准确评估了不同模型在精度与速度上的权衡，避免了试错成本。\n- **数据准备无忧**：迅速获取官方推荐的数据集列表及相关预处理工具，立即启动了模型微调工作。\n\nawesome-human-pose-estimation 将原本碎片化的学术资源转化为结构化的工程弹药库，帮助开发者跳过重复造轮子的阶段，直接站在巨人肩膀上创新。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fcbsudux_awesome-human-pose-estimation_c1c93240.png","cbsudux","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fcbsudux_df15e159.png","Machine Learning Engineer\r\n",null,"Chennai","cbsudu@gmail.com","https:\u002F\u002Fgithub.com\u002Fcbsudux",2481,405,"2026-04-03T05:55:58",5,"","未说明",{"notes":89,"python":87,"dependencies":90},"该仓库是一个资源合集（Awesome List），主要收录了人体姿态估计领域的论文、数据集、博客文章以及指向其他独立项目代码库的链接（涵盖 PyTorch, TensorFlow, Torch 等框架）。它本身不是一个可直接运行的单一软件工具，因此 README 中未包含具体的操作系统、GPU、内存、Python 版本或依赖库的安装要求。用户需根据列表中感兴趣的具体子项目（如 OpenPose, HRNet, Simple Baselines 等）前往其各自的代码仓库查询详细的运行环境需求。",[],[54,14,13],[93,94,95,96,97,98,99,100],"human-pose-estimation","deep-learning","deep-learning-papers","computer-vision","awesome-list","2d-human-pose","3d-human-pose","pose-estimation","2026-03-27T02:49:30.150509","2026-04-06T07:14:25.097339",[104,109],{"id":105,"question_zh":106,"answer_zh":107,"source_url":108},15574,"在哪里可以获取 Human3.6M 数据集的副本？","您可以查看此仓库的数据集：https:\u002F\u002Fgithub.com\u002Fxingyizhou\u002Fpytorch-pose-hg-3d。这是 Human3.6M 的一个采样版本。","https:\u002F\u002Fgithub.com\u002Fcbsudux\u002Fawesome-human-pose-estimation\u002Fissues\u002F5",{"id":110,"question_zh":111,"answer_zh":112,"source_url":113},15575,"是否有基于关节排名（rankings of human joints）来实现更鲁棒结果的 3D 姿态估计算法或论文推荐？","有一篇 IJCAI 2018 的论文提出了利用人体关节排名来实现更鲁棒结果的想法，论文链接为：https:\u002F\u002Fwww.ijcai.org\u002Fproceedings\u002F2018\u002F0136.pdf。此外，相关的 CVPR 2019 论文也已更新收录。","https:\u002F\u002Fgithub.com\u002Fcbsudux\u002Fawesome-human-pose-estimation\u002Fissues\u002F3",[]]