{"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/timeception-for-complex-action-recognition","title":"Timeception for Complex Action Recognition","arxiv_id":"1812.01289","date":"2018-12-04","proceeding":"CVPR 2019 6","authors":["Noureldien Hussein","Efstratios Gavves","Arnold W. M. Smeulders"],"abstract":"This paper focuses on the temporal aspect for recognizing human activities in\nvideos; an important visual cue that has long been undervalued. We revisit the\nconventional definition of activity and restrict it to Complex Action: a set of\none-actions with a weak temporal pattern that serves a specific purpose.\nRelated works use spatiotemporal 3D convolutions with fixed kernel size, too\nrigid to capture the varieties in temporal extents of complex actions, and too\nshort for long-range temporal modeling. In contrast, we use multi-scale\ntemporal convolutions, and we reduce the complexity of 3D convolutions. The\noutcome is Timeception convolution layers, which reasons about minute-long\ntemporal patterns, a factor of 8 longer than best related works. As a result,\nTimeception achieves impressive accuracy in recognizing the human activities of\nCharades, Breakfast Actions, and MultiTHUMOS. Further, we demonstrate that\nTimeception learns long-range temporal dependencies and tolerate temporal\nextents of complex actions.","url_abs":"http://arxiv.org/abs/1812.01289v2","url_pdf":"http://arxiv.org/pdf/1812.01289v2.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":"timeception-for-complex-action-recognition","repo_url":"https://github.com/noureldien/timeception","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"timeception-for-complex-action-recognition","repo_url":"https://github.com/CMU-CREATE-Lab/deep-smoke-machine","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"timeception-for-complex-action-recognition","repo_url":"https://github.com/QUVA-Lab/timeception","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"action-classification","task_name":"Action Classification"},{"task_slug":"action-recognition-in-videos","task_name":"Action Recognition"},{"task_slug":"long-video-activity-recognition","task_name":"Long-video Activity Recognition"},{"task_slug":"video-classification","task_name":"Video Classification"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/action-classification-on-charades","task":"Action Classification","dataset":"Charades","model":"Timeception (R3D)","rank_in_archive_order":30,"of":49,"metrics":{"MAP":"41.1"},"uses_additional_data":false},{"leaderboard":"/sota/action-classification-on-charades","task":"Action Classification","dataset":"Charades","model":"Timeception (I3D)","rank_in_archive_order":38,"of":49,"metrics":{"MAP":"37.2"},"uses_additional_data":false},{"leaderboard":"/sota/action-classification-on-charades","task":"Action Classification","dataset":"Charades","model":"Timeception (R2D)","rank_in_archive_order":41,"of":49,"metrics":{"MAP":"31.6"},"uses_additional_data":false},{"leaderboard":"/sota/long-video-activity-recognition-on-breakfast","task":"Long-video Activity Recognition","dataset":"Breakfast","model":"Timeception (I3D-K400-Pretrain-feature)","rank_in_archive_order":7,"of":8,"metrics":{"mAP":"61.82"},"uses_additional_data":false},{"leaderboard":"/sota/video-classification-on-breakfast","task":"Video Classification","dataset":"Breakfast","model":"Timeception","rank_in_archive_order":8,"of":9,"metrics":{"Accuracy (%)":"71.3"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1812.01289","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}