{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/poseaug-a-differentiable-pose-augmentation","title":"PoseAug: A Differentiable Pose Augmentation Framework for 3D Human Pose Estimation","arxiv_id":"2105.02465","date":"2021-05-06","proceeding":"CVPR 2021 1","authors":["Kehong Gong","Jianfeng Zhang","Jiashi Feng"],"abstract":"Existing 3D human pose estimators suffer poor generalization performance to new datasets, largely due to the limited diversity of 2D-3D pose pairs in the training data. To address this problem, we present PoseAug, a new auto-augmentation framework that learns to augment the available training poses towards a greater diversity and thus improve generalization of the trained 2D-to-3D pose estimator. Specifically, PoseAug introduces a novel pose augmentor that learns to adjust various geometry factors (e.g., posture, body size, view point and position) of a pose through differentiable operations. With such differentiable capacity, the augmentor can be jointly optimized with the 3D pose estimator and take the estimation error as feedback to generate more diverse and harder poses in an online manner. Moreover, PoseAug introduces a novel part-aware Kinematic Chain Space for evaluating local joint-angle plausibility and develops a discriminative module accordingly to ensure the plausibility of the augmented poses. These elaborate designs enable PoseAug to generate more diverse yet plausible poses than existing offline augmentation methods, and thus yield better generalization of the pose estimator. PoseAug is generic and easy to be applied to various 3D pose estimators. Extensive experiments demonstrate that PoseAug brings clear improvements on both intra-scenario and cross-scenario datasets. Notably, it achieves 88.6% 3D PCK on MPI-INF-3DHP under cross-dataset evaluation setup, improving upon the previous best data augmentation based method by 9.1%. Code can be found at: https://github.com/jfzhang95/PoseAug.","url_abs":"https://arxiv.org/abs/2105.02465v1","url_pdf":"https://arxiv.org/pdf/2105.02465v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"poseaug-a-differentiable-pose-augmentation","repo_url":"https://github.com/jfzhang95/PoseAug","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"3d-human-pose-estimation","task_name":"3D Human Pose Estimation"},{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":"diversity","task_name":"Diversity"},{"task_slug":"monocular-3d-human-pose-estimation","task_name":"Monocular 3D Human Pose Estimation"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"},{"task_slug":"weakly-supervised-3d-human-pose-estimation","task_name":"Weakly-supervised 3D Human Pose Estimation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/3d-human-pose-estimation-on-3dpw","task":"3D Human Pose Estimation","dataset":"3DPW","model":"HR-Net+ST-GCN+PoseAug","rank_in_archive_order":107,"of":119,"metrics":{"PA-MPJPE":"73.2"},"uses_additional_data":false},{"leaderboard":"/sota/3d-human-pose-estimation-on-human36m","task":"3D Human Pose Estimation","dataset":"Human3.6M","model":"HR-Net+VPose+PoseAug","rank_in_archive_order":64,"of":88,"metrics":{"Average MPJPE (mm)":"50.2","Multi-View or Monocular":"Monocular","Using 2D ground-truth joints":"No"},"uses_additional_data":false},{"leaderboard":"/sota/3d-human-pose-estimation-on-human36m","task":"3D Human Pose Estimation","dataset":"Human3.6M","model":"HR-Net+ST-GCN+PoseAug","rank_in_archive_order":69,"of":88,"metrics":{"Average MPJPE (mm)":"50.8","Multi-View or Monocular":"Monocular","Using 2D ground-truth joints":"No"},"uses_additional_data":false},{"leaderboard":"/sota/3d-human-pose-estimation-on-mpi-inf-3dhp","task":"3D Human Pose Estimation","dataset":"MPI-INF-3DHP","model":"PoseAug (+Extra2D)","rank_in_archive_order":29,"of":108,"metrics":{"AUC":"57.9","MPJPE":"71.1","PCK":"89.2"},"uses_additional_data":false},{"leaderboard":"/sota/3d-human-pose-estimation-on-mpi-inf-3dhp","task":"3D Human Pose Estimation","dataset":"MPI-INF-3DHP","model":"VPose+PoseAug","rank_in_archive_order":31,"of":108,"metrics":{"AUC":"57.3","MPJPE":"73","PCK":"88.6"},"uses_additional_data":false},{"leaderboard":"/sota/3d-human-pose-estimation-on-mpi-inf-3dhp","task":"3D Human Pose Estimation","dataset":"MPI-INF-3DHP","model":"HR-Net+VPose+PoseAug","rank_in_archive_order":32,"of":108,"metrics":{"MPJPE":"73.2"},"uses_additional_data":false},{"leaderboard":"/sota/3d-human-pose-estimation-on-mpi-inf-3dhp","task":"3D Human Pose Estimation","dataset":"MPI-INF-3DHP","model":"HR-Net+ST-GCN+PoseAug","rank_in_archive_order":35,"of":108,"metrics":{"MPJPE":"76.6"},"uses_additional_data":false},{"leaderboard":"/sota/monocular-3d-human-pose-estimation-on-human3","task":"Monocular 3D Human Pose Estimation","dataset":"Human3.6M","model":"HR-Net+VPose+PoseAug","rank_in_archive_order":23,"of":52,"metrics":{"Average MPJPE (mm)":"50.2","PA-MPJPE":"39.1"},"uses_additional_data":false},{"leaderboard":"/sota/monocular-3d-human-pose-estimation-on-human3","task":"Monocular 3D Human Pose Estimation","dataset":"Human3.6M","model":"PoseAug","rank_in_archive_order":47,"of":52,"metrics":{"Frames Needed":"1","Need Ground Truth 2D Pose":"No","Use Video Sequence":"No"},"uses_additional_data":false},{"leaderboard":"/sota/weakly-supervised-3d-human-pose-estimation-on","task":"Weakly-supervised 3D Human Pose Estimation","dataset":"Human3.6M","model":"PoseAug","rank_in_archive_order":7,"of":33,"metrics":{"3D Annotations":"S1","Average MPJPE (mm)":"56.7","Number of Frames Per View":"1","Number of Views":"1"},"uses_additional_data":false},{"leaderboard":"/sota/weakly-supervised-3d-human-pose-estimation-on","task":"Weakly-supervised 3D Human Pose Estimation","dataset":"Human3.6M","model":"PoseAug","rank_in_archive_order":31,"of":33,"metrics":{"3D Annotations":"S1","Number of Frames Per View":"1"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2105.02465","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}