mmpose/projects/rtmpose3d

RTMPose3D: Real-Time 3D Pose Estimation toolkit based on RTMPose

Technical Report: RTMW: Real-Time Multi-Person 2D and 3D Whole-body Pose Estimation

Abstract

RTMPose3D is a toolkit for real-time 3D pose estimation. It is based on the RTMPose model, which is a 2D pose estimation model that is capable of predicting 2D keypoints and body part associations in real-time. RTMPose3D extends RTMPose by adding a 3D pose estimation branch that can predict 3D keypoints from images directly.

Please refer to our technical report for more details.

🗂️ Model Zoo

Model	AP on COCO-Wholebody	MPJPE on H3WB	Download
RTMW3D-L	0.678	0.056	ckpt
RTMW3D-X	0.680	0.057	ckpt

📚 Usage

👉🏼 TRY RTMPose3D NOW

cd /path/to/mmpose/projects/rtmpose3d
export PYTHONPATH=$(pwd):$PYTHONPATH
python body3d_img2pose_demo.py configs/rtmdet_m_640-8xb32_coco-person.py https://download.openmmlab.com/mmpose/v1/projects/rtmpose/rtmdet_m_8xb32-100e_coco-obj365-person-235e8209.pth configs\rtmw3d-l_8xb64_cocktail14-384x288.py rtmw3d-l_cock14-0d4ad840_20240422.pth --input /path/to/image --output-root /path/to/output

📜 Citation 🔝

If you find RTMPose3D toolkit or RTMW3D models useful in your research, please consider cite:

@article{jiang2024rtmw,
  title={RTMW: Real-Time Multi-Person 2D and 3D Whole-body Pose Estimation},
  author={Jiang, Tao and Xie, Xinchen and Li, Yining},
  journal={arXiv preprint arXiv:2407.08634},
  year={2024}
}

@misc{mmpose2020,
    title={OpenMMLab Pose Estimation Toolbox and Benchmark},
    author={MMPose Contributors},
    howpublished = {\url{https://github.com/open-mmlab/mmpose}},
    year={2020}
}