{"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/posetrack-joint-multi-person-pose-estimation","title":"PoseTrack: Joint Multi-Person Pose Estimation and Tracking","arxiv_id":"1611.07727","date":"2016-11-23","proceeding":"CVPR 2017 7","authors":["Umar Iqbal","Anton Milan","Juergen Gall"],"abstract":"In this work, we introduce the challenging problem of joint multi-person pose\nestimation and tracking of an unknown number of persons in unconstrained\nvideos. Existing methods for multi-person pose estimation in images cannot be\napplied directly to this problem, since it also requires to solve the problem\nof person association over time in addition to the pose estimation for each\nperson. We therefore propose a novel method that jointly models multi-person\npose estimation and tracking in a single formulation. To this end, we represent\nbody joint detections in a video by a spatio-temporal graph and solve an\ninteger linear program to partition the graph into sub-graphs that correspond\nto plausible body pose trajectories for each person. The proposed approach\nimplicitly handles occlusion and truncation of persons. Since the problem has\nnot been addressed quantitatively in the literature, we introduce a challenging\n\"Multi-Person PoseTrack\" dataset, and also propose a completely unconstrained\nevaluation protocol that does not make any assumptions about the scale, size,\nlocation or the number of persons. Finally, we evaluate the proposed approach\nand several baseline methods on our new dataset.","url_abs":"http://arxiv.org/abs/1611.07727v3","url_pdf":"http://arxiv.org/pdf/1611.07727v3.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":"posetrack-joint-multi-person-pose-estimation","repo_url":"https://github.com/iqbalu/PoseTrack-CVPR2017","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"posetrack-joint-multi-person-pose-estimation","repo_url":"https://github.com/umariqb/posetrack-cvpr2017","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"multi-person-pose-estimation","task_name":"Multi-Person Pose Estimation"},{"task_slug":"multi-person-pose-estimation-and-tracking","task_name":"Multi-Person Pose Estimation and Tracking"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"},{"task_slug":"pose-tracking","task_name":"Pose Tracking"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/multi-person-pose-estimation-on-multi-person","task":"Multi-Person Pose Estimation","dataset":"Multi-Person PoseTrack","model":"PoseTrack","rank_in_archive_order":1,"of":1,"metrics":{"Mean mAP":"38.2"},"uses_additional_data":false},{"leaderboard":"/sota/pose-tracking-on-multi-person-posetrack","task":"Pose Tracking","dataset":"Multi-Person PoseTrack","model":"PoseTrack","rank_in_archive_order":1,"of":1,"metrics":{"MOTA":"28.2","MOTP":"55.7"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1611.07727","atlas_url":"https://app.syntology.ai/?focus=1611.07727","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}