-
[pdf]
[supp]
[arXiv]
[bibtex]@InProceedings{Li_2021_ICCV, author = {Li, Yanjie and Zhang, Shoukui and Wang, Zhicheng and Yang, Sen and Yang, Wankou and Xia, Shu-Tao and Zhou, Erjin}, title = {TokenPose: Learning Keypoint Tokens for Human Pose Estimation}, booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)}, month = {October}, year = {2021}, pages = {11313-11322} }
TokenPose: Learning Keypoint Tokens for Human Pose Estimation
Abstract
Human pose estimation deeply relies on visual clues and anatomical constraints between parts to locate keypoints. Most existing CNN-based methods do well in visual representation, however, lacking in the ability to explicitly learn the constraint relationships between keypoints. In this paper, we propose a novel approach based on Token representation for human Pose estimation (TokenPose). In detail, each keypoint is explicitly embedded as a token to simultaneously learn constraint relationships and appearance cues from images. Extensive experiments show that the small and large TokenPose models are on par with state-of-the-art CNN-based counterparts while being more lightweight. Specifically, our TokenPose-S and TokenPose-L achieve 72.5 AP and 75.8 AP on COCO validation dataset respectively, with significant reduction in parameters and GFLOPs. Code is publicly available at https://github.com/leeyegy/TokenPose.
Related Material