{"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/keep-it-smpl-automatic-estimation-of-3d-human","title":"Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image","arxiv_id":"1607.08128","date":"2016-07-27","proceeding":null,"authors":["Federica Bogo","Angjoo Kanazawa","Christoph Lassner","Peter Gehler","Javier Romero","Michael J. Black"],"abstract":"We describe the first method to automatically estimate the 3D pose of the\nhuman body as well as its 3D shape from a single unconstrained image. We\nestimate a full 3D mesh and show that 2D joints alone carry a surprising amount\nof information about body shape. The problem is challenging because of the\ncomplexity of the human body, articulation, occlusion, clothing, lighting, and\nthe inherent ambiguity in inferring 3D from 2D. To solve this, we first use a\nrecently published CNN-based method, DeepCut, to predict (bottom-up) the 2D\nbody joint locations. We then fit (top-down) a recently published statistical\nbody shape model, called SMPL, to the 2D joints. We do so by minimizing an\nobjective function that penalizes the error between the projected 3D model\njoints and detected 2D joints. Because SMPL captures correlations in human\nshape across the population, we are able to robustly fit it to very little\ndata. We further leverage the 3D model to prevent solutions that cause\ninterpenetration. We evaluate our method, SMPLify, on the Leeds Sports,\nHumanEva, and Human3.6M datasets, showing superior pose accuracy with respect\nto the state of the art.","url_abs":"http://arxiv.org/abs/1607.08128v1","url_pdf":"http://arxiv.org/pdf/1607.08128v1.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":"keep-it-smpl-automatic-estimation-of-3d-human","repo_url":"https://github.com/Jtoo/fitting_human_smpl_model","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"keep-it-smpl-automatic-estimation-of-3d-human","repo_url":"https://github.com/dizhongzhu/Humanbody_Literatures","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"3d-human-pose-estimation","task_name":"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":"SMPLify","rank_in_archive_order":112,"of":119,"metrics":{"PA-MPJPE":"106.8"},"uses_additional_data":false},{"leaderboard":"/sota/3d-human-pose-estimation-on-humaneva-i","task":"3D Human Pose Estimation","dataset":"HumanEva-I","model":"SMPLify","rank_in_archive_order":31,"of":31,"metrics":{"Mean Reconstruction Error (mm)":"79.9"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1607.08128","atlas_url":"https://app.syntology.ai/?focus=1607.08128","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}