{"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/posefix-model-agnostic-general-human-pose","title":"PoseFix: Model-agnostic General Human Pose Refinement Network","arxiv_id":"1812.03595","date":"2018-12-10","proceeding":"CVPR 2019 6","authors":["Gyeongsik Moon","Ju Yong Chang","Kyoung Mu Lee"],"abstract":"Multi-person pose estimation from a 2D image is an essential technique for\nhuman behavior understanding. In this paper, we propose a human pose refinement\nnetwork that estimates a refined pose from a tuple of an input image and input\npose. The pose refinement was performed mainly through an end-to-end trainable\nmulti-stage architecture in previous methods. However, they are highly\ndependent on pose estimation models and require careful model design. By\ncontrast, we propose a model-agnostic pose refinement method. According to a\nrecent study, state-of-the-art 2D human pose estimation methods have similar\nerror distributions. We use this error statistics as prior information to\ngenerate synthetic poses and use the synthesized poses to train our model. In\nthe testing stage, pose estimation results of any other methods can be input to\nthe proposed method. Moreover, the proposed model does not require code or\nknowledge about other methods, which allows it to be easily used in the\npost-processing step. We show that the proposed approach achieves better\nperformance than the conventional multi-stage refinement models and\nconsistently improves the performance of various state-of-the-art pose\nestimation methods on the commonly used benchmark. The code is available in\nthis https URL\\footnote{\\url{https://github.com/mks0601/PoseFix_RELEASE}}.","url_abs":"http://arxiv.org/abs/1812.03595v3","url_pdf":"http://arxiv.org/pdf/1812.03595v3.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":"posefix-model-agnostic-general-human-pose","repo_url":"https://github.com/mks0601/PoseFix_RELEASE","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"2d-human-pose-estimation","task_name":"2D Human Pose Estimation"},{"task_slug":"keypoint-detection","task_name":"Keypoint Detection"},{"task_slug":"multi-person-pose-estimation","task_name":"Multi-Person Pose Estimation"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"},{"task_slug":"model","task_name":"model"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/keypoint-detection-on-coco","task":"Keypoint Detection","dataset":"COCO (Common Objects in Context)","model":"PoseFix(384x288)","rank_in_archive_order":4,"of":24,"metrics":{"Test AP":"76.7","Validation AP":"77.3"},"uses_additional_data":false},{"leaderboard":"/sota/multi-person-pose-estimation-on-coco","task":"Multi-Person Pose Estimation","dataset":"COCO (Common Objects in Context)","model":"PoseFix","rank_in_archive_order":14,"of":15,"metrics":{"Test AP":"76.7","Validation AP":"77.3"},"uses_additional_data":false},{"leaderboard":"/sota/pose-estimation-on-coco-test-dev","task":"Pose Estimation","dataset":"COCO test-dev","model":"PoseFix","rank_in_archive_order":22,"of":47,"metrics":{"AP":"74.7","AP50":"91.2","AP75":"81.9","APL":"81.2","APM":"71.1","AR":"79.9"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1812.03595","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}