{"id":15034007,"url":"https://github.com/mvig-sjtu/alphapose","last_synced_at":"2025-04-25T14:45:55.819Z","repository":{"id":37359709,"uuid":"117772329","full_name":"MVIG-SJTU/AlphaPose","owner":"MVIG-SJTU","description":"Real-Time and Accurate Full-Body Multi-Person Pose Estimation\u0026Tracking System","archived":false,"fork":false,"pushed_at":"2024-05-13T14:33:34.000Z","size":121674,"stargazers_count":8212,"open_issues_count":300,"forks_count":1991,"subscribers_count":208,"default_branch":"master","last_synced_at":"2025-04-09T00:28:31.283Z","etag":null,"topics":["accurate","alpha-pose","alphapose","crowdpose","full-body","gpu","human-computer-interaction","human-joints","human-pose-estimation","human-pose-tracking","human-tracking","keypoints","person-pose-estimation","pose-estimation","posetracking","pytorch","realtime","skeleton","tracking","whole-body"],"latest_commit_sha":null,"homepage":"http://mvig.org/research/alphapose.html","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/MVIG-SJTU.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2018-01-17T02:33:17.000Z","updated_at":"2025-04-08T03:11:18.000Z","dependencies_parsed_at":"2023-02-09T13:15:44.347Z","dependency_job_id":"c730fa6b-1fb1-4086-8a03-e8af507d47a7","html_url":"https://github.com/MVIG-SJTU/AlphaPose","commit_stats":{"total_commits":175,"total_committers":19,"mean_commits":9.210526315789474,"dds":0.4685714285714285,"last_synced_commit":"c60106d19afb443e964df6f06ed1842962f5f1f7"},"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MVIG-SJTU%2FAlphaPose","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MVIG-SJTU%2FAlphaPose/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MVIG-SJTU%2FAlphaPose/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MVIG-SJTU%2FAlphaPose/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/MVIG-SJTU","download_url":"https://codeload.github.com/MVIG-SJTU/AlphaPose/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":250560986,"owners_count":21450332,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["accurate","alpha-pose","alphapose","crowdpose","full-body","gpu","human-computer-interaction","human-joints","human-pose-estimation","human-pose-tracking","human-tracking","keypoints","person-pose-estimation","pose-estimation","posetracking","pytorch","realtime","skeleton","tracking","whole-body"],"created_at":"2024-09-24T20:23:35.940Z","updated_at":"2025-04-24T04:20:24.529Z","avatar_url":"https://github.com/MVIG-SJTU.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n\u003cdiv align=\"center\"\u003e\n    \u003cimg src=\"docs/logo.jpg\", width=\"400\"\u003e\n\u003c/div\u003e\n\n\n## News!\n- Nov 2022: [**AlphaPose paper**](http://arxiv.org/abs/2211.03375) is released! Checkout the paper for more details about this project.\n- Sep 2022: [**Jittor** version](https://github.com/tycoer/AlphaPose_jittor) of AlphaPose is released! It achieves 1.45x speed up with resnet50 backbone on the training stage.\n- July 2022: [**v0.6.0** version](https://github.com/MVIG-SJTU/AlphaPose) of AlphaPose is released! [HybrIK](https://github.com/Jeff-sjtu/HybrIK) for 3D pose and shape estimation is supported!\n- Jan 2022: [**v0.5.0** version](https://github.com/MVIG-SJTU/AlphaPose) of AlphaPose is released! Stronger whole body(face,hand,foot) keypoints! More models are availabel. Checkout [docs/MODEL_ZOO.md](docs/MODEL_ZOO.md)\n- Aug 2020: [**v0.4.0** version](https://github.com/MVIG-SJTU/AlphaPose) of AlphaPose is released! Stronger tracking! Include whole body(face,hand,foot) keypoints! [Colab](https://colab.research.google.com/drive/1c7xb_7U61HmeJp55xjXs24hf1GUtHmPs?usp=sharing) now available.\n- Dec 2019: [**v0.3.0** version](https://github.com/MVIG-SJTU/AlphaPose) of AlphaPose is released! Smaller model, higher accuracy!\n- Apr 2019: [**MXNet** version](https://github.com/MVIG-SJTU/AlphaPose/tree/mxnet) of AlphaPose is released! It runs at **23 fps** on COCO validation set.\n- Feb 2019: [CrowdPose](https://github.com/MVIG-SJTU/AlphaPose/docs/CrowdPose.md) is integrated into AlphaPose Now!\n- Dec 2018: [General version](https://github.com/MVIG-SJTU/AlphaPose/trackers/PoseFlow) of PoseFlow is released! 3X Faster and support pose tracking results visualization!\n- Sep 2018: [**v0.2.0** version](https://github.com/MVIG-SJTU/AlphaPose/tree/pytorch) of AlphaPose is released! It runs at **20 fps** on COCO validation set (4.6 people per image on average) and achieves 71 mAP!\n\n## AlphaPose\n[AlphaPose](http://www.mvig.org/research/alphapose.html) is an accurate multi-person pose estimator, which is the **first open-source system that achieves 70+ mAP (75 mAP) on COCO dataset and 80+ mAP (82.1 mAP) on MPII dataset.** \nTo match poses that correspond to the same person across frames, we also provide an efficient online pose tracker called Pose Flow. It is the **first open-source online pose tracker that achieves both 60+ mAP (66.5 mAP) and 50+ MOTA (58.3 MOTA) on PoseTrack Challenge dataset.**\n\nAlphaPose supports both Linux and **Windows!**\n\n\u003cdiv align=\"center\"\u003e\n    \u003cimg src=\"docs/alphapose_17.gif\", width=\"400\" alt\u003e\u003cbr\u003e\n    COCO 17 keypoints\n\u003c/div\u003e\n\u003cdiv align=\"center\"\u003e\n    \u003cimg src=\"docs/alphapose_26.gif\", width=\"400\" alt\u003e\u003cbr\u003e\n    \u003cb\u003e\u003ca href=\"https://github.com/Fang-Haoshu/Halpe-FullBody\"\u003eHalpe 26 keypoints\u003c/a\u003e\u003c/b\u003e + tracking\n\u003c/div\u003e\n\u003cdiv align=\"center\"\u003e\n    \u003cimg src=\"docs/alphapose_136.gif\", width=\"400\"alt\u003e\u003cbr\u003e\n    \u003cb\u003e\u003ca href=\"https://github.com/Fang-Haoshu/Halpe-FullBody\"\u003eHalpe 136 keypoints\u003c/a\u003e\u003c/b\u003e + tracking\n    \u003cb\u003e\u003ca href=\"https://youtu.be/uze6chg-YeU\"\u003eYouTube link\u003c/a\u003e\u003c/b\u003e\u003cbr\u003e\n\u003c/div\u003e\n\u003cdiv align=\"center\"\u003e\n    \u003cimg src=\"docs/alphapose_hybrik_smpl.gif\", width=\"400\"alt\u003e\u003cbr\u003e\n    \u003cb\u003e\u003ca href=\"https://github.com/Jeff-sjtu/HybrIK\"\u003eSMPL\u003c/a\u003e\u003c/b\u003e + tracking\n\u003c/div\u003e\n\n\n## Results\n### Pose Estimation\nResults on COCO test-dev 2015:\n\u003ccenter\u003e\n\n| Method | AP @0.5:0.95 | AP @0.5 | AP @0.75 | AP medium | AP large |\n|:-------|:-----:|:-------:|:-------:|:-------:|:-------:|\n| OpenPose (CMU-Pose) | 61.8 | 84.9 | 67.5 | 57.1 | 68.2 |\n| Detectron (Mask R-CNN) | 67.0 | 88.0 | 73.1 | 62.2 | 75.6 |\n| **AlphaPose** | **73.3** | **89.2** | **79.1** | **69.0** | **78.6** |\n\n\u003c/center\u003e\n\nResults on MPII full test set:\n\u003ccenter\u003e\n\n| Method | Head | Shoulder | Elbow | Wrist | Hip | Knee | Ankle | Ave |\n|:-------|:-----:|:-------:|:-------:|:-------:|:-------:|:-------:|:-------:|:-------:|\n| OpenPose (CMU-Pose) | 91.2 | 87.6 | 77.7 | 66.8 | 75.4 | 68.9 | 61.7 | 75.6 |\n| Newell \u0026 Deng | **92.1** | 89.3 | 78.9 | 69.8 | 76.2 | 71.6 | 64.7 | 77.5 |\n| **AlphaPose** | 91.3 | **90.5** | **84.0** | **76.4** | **80.3** | **79.9** | **72.4** | **82.1** |\n\n\u003c/center\u003e\n\nMore results and models are available in the [docs/MODEL_ZOO.md](docs/MODEL_ZOO.md).\n\n### Pose Tracking\n\n\u003cp align='center'\u003e\n    \u003cimg src=\"docs/posetrack.gif\", width=\"360\"\u003e\n    \u003cimg src=\"docs/posetrack2.gif\", width=\"344\"\u003e\n\u003c/p\u003e\n\nPlease read [trackers/README.md](trackers/) for details.\n\n### CrowdPose\n\u003cp align='center'\u003e\n    \u003cimg src=\"docs/crowdpose.gif\", width=\"360\"\u003e\n\u003c/p\u003e\n\nPlease read [docs/CrowdPose.md](docs/CrowdPose.md) for details.\n\n\n## Installation\nPlease check out [docs/INSTALL.md](docs/INSTALL.md)\n\n## Model Zoo\nPlease check out [docs/MODEL_ZOO.md](docs/MODEL_ZOO.md)\n\n## Quick Start\n- **Colab**: We provide a [colab example](https://colab.research.google.com/drive/1_3Wxi4H3QGVC28snL3rHIoeMAwI2otMR?usp=sharing) for your quick start.\n\n- **Inference**: Inference demo\n``` bash\n./scripts/inference.sh ${CONFIG} ${CHECKPOINT} ${VIDEO_NAME} # ${OUTPUT_DIR}, optional\n```\n\nInference SMPL (Download the SMPL model `basicModel_neutral_lbs_10_207_0_v1.0.0.pkl` from [here](https://smpl.is.tue.mpg.de/) and put it in `model_files/`).\n``` bash\n./scripts/inference_3d.sh ./configs/smpl/256x192_adam_lr1e-3-res34_smpl_24_3d_base_2x_mix.yaml ${CHECKPOINT} ${VIDEO_NAME} # ${OUTPUT_DIR}, optional\n```\nFor high level API, please refer to `./scripts/demo_api.py`. To enable tracking, please refer to [this page](./trackers).\n\n- **Training**: Train from scratch\n``` bash\n./scripts/train.sh ${CONFIG} ${EXP_ID}\n```\n\n- **Validation**: Validate your model on MSCOCO val2017\n``` bash\n./scripts/validate.sh ${CONFIG} ${CHECKPOINT}\n```\n\nExamples:\n\nDemo using `FastPose` model.\n``` bash\n./scripts/inference.sh configs/coco/resnet/256x192_res50_lr1e-3_1x.yaml pretrained_models/fast_res50_256x192.pth ${VIDEO_NAME}\n#or\npython scripts/demo_inference.py --cfg configs/coco/resnet/256x192_res50_lr1e-3_1x.yaml --checkpoint pretrained_models/fast_res50_256x192.pth --indir examples/demo/\n#or if you want to use yolox-x as the detector\npython scripts/demo_inference.py --detector yolox-x --cfg configs/coco/resnet/256x192_res50_lr1e-3_1x.yaml --checkpoint pretrained_models/fast_res50_256x192.pth --indir examples/demo/\n```\n\nTrain `FastPose` on mscoco dataset.\n``` bash\n./scripts/train.sh ./configs/coco/resnet/256x192_res50_lr1e-3_1x.yaml exp_fastpose\n```\n\nMore detailed inference options and examples, please refer to [GETTING_STARTED.md](docs/GETTING_STARTED.md)\n\n\n## Common issue \u0026 FAQ\nCheck out [faq.md](docs/faq.md) for faq. If it can not solve your problems or if you find any bugs, don't hesitate to comment on GitHub or make a pull request!\n\n## Contributors\nAlphaPose is based on RMPE(ICCV'17), authored by [Hao-Shu Fang](https://fang-haoshu.github.io/), Shuqin Xie, [Yu-Wing Tai](https://scholar.google.com/citations?user=nFhLmFkAAAAJ\u0026hl=en) and [Cewu Lu](http://www.mvig.org/), [Cewu Lu](http://mvig.sjtu.edu.cn/) is the corresponding author. Currently, it is maintained by [Jiefeng Li\\*](http://jeff-leaf.site/), [Hao-shu Fang\\*](https://fang-haoshu.github.io/),  [Haoyi Zhu](https://github.com/HaoyiZhu), [Yuliang Xiu](http://xiuyuliang.cn/about/) and [Chao Xu](http://www.isdas.cn/). \n\nThe main contributors are listed in [doc/contributors.md](docs/contributors.md).\n\n## TODO\n- [x] Multi-GPU/CPU inference\n- [x] 3D pose\n- [x] add tracking flag\n- [ ] PyTorch C++ version\n- [x] Add model trained on mixture dataset (Check the model zoo)\n- [ ] dense support\n- [x] small box easy filter\n- [x] Crowdpose support\n- [ ] Speed up PoseFlow\n- [x] Add stronger/light detectors (yolox is now supported)\n- [x] High level API (check the scripts/demo_api.py)\n\nWe would really appreciate if you can offer any help and be the [contributor](docs/contributors.md) of AlphaPose.\n\n\n## Citation\nPlease cite these papers in your publications if it helps your research:\n\n    @article{alphapose,\n      author = {Fang, Hao-Shu and Li, Jiefeng and Tang, Hongyang and Xu, Chao and Zhu, Haoyi and Xiu, Yuliang and Li, Yong-Lu and Lu, Cewu},\n      journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},\n      title = {AlphaPose: Whole-Body Regional Multi-Person Pose Estimation and Tracking in Real-Time},\n      year = {2022}\n    }\n    \n    @inproceedings{fang2017rmpe,\n      title={{RMPE}: Regional Multi-person Pose Estimation},\n      author={Fang, Hao-Shu and Xie, Shuqin and Tai, Yu-Wing and Lu, Cewu},\n      booktitle={ICCV},\n      year={2017}\n    }\n\n    @inproceedings{li2019crowdpose,\n        title={Crowdpose: Efficient crowded scenes pose estimation and a new benchmark},\n        author={Li, Jiefeng and Wang, Can and Zhu, Hao and Mao, Yihuan and Fang, Hao-Shu and Lu, Cewu},\n        booktitle={Proceedings of the IEEE/CVF conference on computer vision and pattern recognition},\n        pages={10863--10872},\n        year={2019}\n    }\n\nIf you used the 3D mesh reconstruction module, please also cite:\n\n    @inproceedings{li2021hybrik,\n        title={Hybrik: A hybrid analytical-neural inverse kinematics solution for 3d human pose and shape estimation},\n        author={Li, Jiefeng and Xu, Chao and Chen, Zhicun and Bian, Siyuan and Yang, Lixin and Lu, Cewu},\n        booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},\n        pages={3383--3393},\n        year={2021}\n    }\n\nIf you used the PoseFlow tracking module, please also cite:\n\n    @inproceedings{xiu2018poseflow,\n      author = {Xiu, Yuliang and Li, Jiefeng and Wang, Haoyu and Fang, Yinghong and Lu, Cewu},\n      title = {{Pose Flow}: Efficient Online Pose Tracking},\n      booktitle={BMVC},\n      year = {2018}\n    }\n\n\n\n\n\n## License\nAlphaPose is freely available for free non-commercial use, and may be redistributed under these conditions. For commercial queries, please drop an e-mail at mvig.alphapose[at]gmail[dot]com and cc lucewu[[at]sjtu[dot]edu[dot]cn. We will send the detail agreement to you.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmvig-sjtu%2Falphapose","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmvig-sjtu%2Falphapose","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmvig-sjtu%2Falphapose/lists"}