CINXE.COM
CVPR 2022 Open Access Repository
<!DOCTYPE html> <html lang="en"> <head> <meta content="text/html; charset=UTF-8" http-equiv="content-type"> <meta name="viewport" content="width=device-width, initial-scale=1"> <link rel="icon" type="image/png" href="/favicon.ico"> <title>CVPR 2022 Open Access Repository</title> <link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.min.css"> <script type="text/javascript" src="https://ajax.googleapis.com/ajax/libs/jquery/3.1.1/jquery.min.js"></script> <script type="text/javascript" src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js"></script> <script type="text/javascript" src="/static/jquery.js"></script> <link rel="stylesheet" type="text/css" href="/static/conf.css"> <meta name="citation_title" content="Revisiting Skeleton-Based Action Recognition"> <meta name="citation_author" content="Duan, Haodong"> <meta name="citation_author" content="Zhao, Yue"> <meta name="citation_author" content="Chen, Kai"> <meta name="citation_author" content="Lin, Dahua"> <meta name="citation_author" content="Dai, Bo"> <meta name="citation_publication_date" content="2022"> <meta name="citation_conference_title" content="Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition"> <meta name="citation_firstpage" content="2969"> <meta name="citation_lastpage" content="2978"> <meta name="citation_pdf_url" content="https://openaccess.thecvf.com/content/CVPR2022/papers/Duan_Revisiting_Skeleton-Based_Action_Recognition_CVPR_2022_paper.pdf"> </head> <body> <div id="header"> <div id="header_left"> <a href="https://cvpr2022.thecvf.com/"><img src="/img/cvpr2022.png" width="175" border="0" alt="CVPR 2022"></a> <a href="https://www.thecvf.com/"><img src="/img/cropped-cvf-s.jpg" width="175" height="112" border="0" alt="CVF"></a> </div> <div id="header_right"> <div id="header_title"> <a href="https://cvpr2022.thecvf.com/">CVPR 2022</a> <a href="/menu" class="a_monochrome">open access</a> </div> <div id="help"> These CVPR 2022 papers are the Open Access versions, provided by the <a href="https://www.thecvf.com/">Computer Vision Foundation.</a><br> Except for the watermark, they are identical to the accepted versions; the final published version of the proceedings is available on IEEE Xplore. </div> <div id="disclaimer"> This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright.<br><br> <form action="/CVPR2022" method="post"> <input type="text" name="query"> <input type="submit" value="Search"> </form> </div> </div> <div id="header_sponsor"> <p style="vertical-align:center; text-align: center"> <strong>Powered by:</strong></p> <img src="/img/ms-azure-logo.png" width="100" alt="Microsoft Azure"> <p> </p> <p> </p> <p style="vertical-align:center; text-align: center"> <strong>Sponsored by:</strong></p> <img src="/img/amazon-logo.png" width="100" alt="Amazon"> <img src="/img/facebook_logo.jpg" width="100" alt="Facebook"> <img src="/img/Google_2015_logo.svg" width="100" alt="Google"> </div> </div> <div class="clear"></div> <div id="content"> <dl> <div id="papertitle"> Revisiting Skeleton-Based Action Recognition <dd> </div> <div id="authors"> <br><b><i>Haodong Duan, Yue Zhao, Kai Chen, Dahua Lin, Bo Dai</i></b>; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 2969-2978 </div> <font size="5"> <br><b>Abstract</b> </font> <br><br> <div id="abstract"> Human skeleton, as a compact representation of human action, has received increasing attention in recent years. Many skeleton-based action recognition methods adopt GCNs to extract features on top of human skeletons. Despite the positive results shown in these attempts, GCN-based methods are subject to limitations in robustness, interoperability, and scalability. In this work, we propose PoseConv3D, a new approach to skeleton-based action recognition. PoseConv3D relies on a 3D heatmap volume instead of a graph sequence as the base representation of human skeletons. Compared to GCN-based methods, PoseConv3D is more effective in learning spatiotemporal features, more robust against pose estimation noises, and generalizes better in cross-dataset settings. Also, PoseConv3D can handle multiple-person scenarios without additional computation costs. The hierarchical features can be easily integrated with other modalities at early fusion stages, providing a great design space to boost the performance. PoseConv3D achieves the state-of-the-art on five of six standard skeleton-based action recognition benchmarks. Once fused with other modalities, it achieves the state-of-the-art on all eight multi-modality action recognition benchmarks. Code has been made available at: https://github.com/kennymckormick/pyskl. </div> <font size="5"> <br><b>Related Material</b> </font> <br><br> <dd> [<a href="/content/CVPR2022/papers/Duan_Revisiting_Skeleton-Based_Action_Recognition_CVPR_2022_paper.pdf">pdf</a>] [<a href="/content/CVPR2022/supplemental/Duan_Revisiting_Skeleton-Based_Action_CVPR_2022_supplemental.zip">supp</a>] [<a href="http://arxiv.org/abs/2104.13586">arXiv</a>] <div class="link2">[<a class="fakelink" onclick="$(this).siblings('.bibref').slideToggle()">bibtex</a>] <div class="bibref pre-white-space">@InProceedings{Duan_2022_CVPR, author = {Duan, Haodong and Zhao, Yue and Chen, Kai and Lin, Dahua and Dai, Bo}, title = {Revisiting Skeleton-Based Action Recognition}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2022}, pages = {2969-2978} }</div> </div> </dd></dl></div> </body> </html>