{"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/fast-video-shot-transition-localization-with","title":"Fast Video Shot Transition Localization with Deep Structured Models","arxiv_id":"1808.04234","date":"2018-08-13","proceeding":null,"authors":["Shitao Tang","Litong Feng","Zhangkui Kuang","Yimin Chen","Wei zhang"],"abstract":"Detection of video shot transition is a crucial pre-processing step in video\nanalysis. Previous studies are restricted on detecting sudden content changes\nbetween frames through similarity measurement and multi-scale operations are\nwidely utilized to deal with transitions of various lengths. However,\nlocalization of gradual transitions are still under-explored due to the high\nvisual similarity between adjacent frames. Cut shot transitions are abrupt\nsemantic breaks while gradual shot transitions contain low-level\nspatial-temporal patterns caused by video effects in addition to the gradual\nsemantic breaks, e.g. dissolve. In order to address the problem, we propose a\nstructured network which is able to detect these two shot transitions using\ntargeted models separately. Considering speed performance trade-offs, we design\na smart framework. With one TITAN GPU, the proposed method can achieve a\n30\\(\\times\\) real-time speed. Experiments on public TRECVID07 and RAI databases\nshow that our method outperforms the state-of-the-art methods. In order to\ntrain a high-performance shot transition detector, we contribute a new database\nClipShots, which contains 128636 cut transitions and 38120 gradual transitions\nfrom 4039 online videos. ClipShots intentionally collect short videos for more\nhard cases caused by hand-held camera vibrations, large object motions, and\nocclusion.","url_abs":"http://arxiv.org/abs/1808.04234v1","url_pdf":"http://arxiv.org/pdf/1808.04234v1.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":"fast-video-shot-transition-localization-with","repo_url":"https://github.com/Tangshitao/ClipShots","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"fast-video-shot-transition-localization-with","repo_url":"https://github.com/Tangshitao/ClipShots_basline","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"fast-video-shot-transition-localization-with","repo_url":"https://github.com/soCzech/TransNetV2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"fast-video-shot-transition-localization-with","repo_url":"https://github.com/wqliu657/TransNetV2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"camera-shot-boundary-detection","task_name":"Camera shot boundary detection"},{"task_slug":null,"task_name":"GPU"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[{"slug":"clipshots","name":"ClipShots","full_name":null}],"methods_introduced":[],"results":[{"leaderboard":"/sota/camera-shot-boundary-detection-on-clipshots","task":"Camera shot boundary detection","dataset":"ClipShots","model":"DSM Cut transition detector","rank_in_archive_order":3,"of":4,"metrics":{"F1 score":"76.1"},"uses_additional_data":true}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1808.04234","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}