{"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/gaitset-regarding-gait-as-a-set-for-cross","title":"GaitSet: Regarding Gait as a Set for Cross-View Gait Recognition","arxiv_id":"1811.06186","date":"2018-11-15","proceeding":null,"authors":["Hanqing Chao","Yiwei He","Junping Zhang","Jianfeng Feng"],"abstract":"As a unique biometric feature that can be recognized at a distance, gait has\nbroad applications in crime prevention, forensic identification and social\nsecurity. To portray a gait, existing gait recognition methods utilize either a\ngait template, where temporal information is hard to preserve, or a gait\nsequence, which must keep unnecessary sequential constraints and thus loses the\nflexibility of gait recognition. In this paper we present a novel perspective,\nwhere a gait is regarded as a set consisting of independent frames. We propose\na new network named GaitSet to learn identity information from the set. Based\non the set perspective, our method is immune to permutation of frames, and can\nnaturally integrate frames from different videos which have been filmed under\ndifferent scenarios, such as diverse viewing angles, different clothes/carrying\nconditions. Experiments show that under normal walking conditions, our\nsingle-model method achieves an average rank-1 accuracy of 95.0% on the CASIA-B\ngait dataset and an 87.1% accuracy on the OU-MVLP gait dataset. These results\nrepresent new state-of-the-art recognition accuracy. On various complex\nscenarios, our model exhibits a significant level of robustness. It achieves\naccuracies of 87.2% and 70.4% on CASIA-B under bag-carrying and coat-wearing\nwalking conditions, respectively. These outperform the existing best methods by\na large margin. The method presented can also achieve a satisfactory accuracy\nwith a small number of frames in a test sample, e.g., 82.5% on CASIA-B with\nonly 7 frames. The source code has been released at\nhttps://github.com/AbnerHqC/GaitSet.","url_abs":"http://arxiv.org/abs/1811.06186v4","url_pdf":"http://arxiv.org/pdf/1811.06186v4.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":"gaitset-regarding-gait-as-a-set-for-cross","repo_url":"https://github.com/AbnerHqC/GaitSet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"gaitset-regarding-gait-as-a-set-for-cross","repo_url":"https://github.com/Gait3D/Gait3D-Benchmark","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"gaitset-regarding-gait-as-a-set-for-cross","repo_url":"https://github.com/shiqiyu/opengait","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"gait-recognition","task_name":"Gait Recognition"},{"task_slug":"multiview-gait-recognition","task_name":"Multiview Gait Recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/gait-recognition-on-oumvlp","task":"Gait Recognition","dataset":"OUMVLP","model":"GaitSet","rank_in_archive_order":7,"of":7,"metrics":{"Averaged rank-1 acc(%)":"87.1"},"uses_additional_data":false},{"leaderboard":"/sota/multiview-gait-recognition-on-casia-b","task":"Multiview Gait Recognition","dataset":"CASIA-B","model":"GaitSet","rank_in_archive_order":10,"of":12,"metrics":{"Accuracy (Cross-View, Avg)":"84.2","BG#1-2":"87.2","CL#1-2":"70.4","NM#5-6 ":"95.0"},"uses_additional_data":false},{"leaderboard":"/sota/multiview-gait-recognition-on-ou-mvlp","task":"Multiview Gait Recognition","dataset":"OU-MVLP","model":"GaitSet","rank_in_archive_order":2,"of":2,"metrics":{"Accuracy (Cross-View)":"87.1"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1811.06186","atlas_url":"https://app.syntology.ai/?focus=1811.06186","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}