{"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/tracking-for-half-an-hour","title":"Tracking for Half an Hour","arxiv_id":"1711.10217","date":"2017-11-28","proceeding":null,"authors":["Ran Tao","Efstratios Gavves","Arnold W. M. Smeulders"],"abstract":"Long-term tracking requires extreme stability to the multitude of model\nupdates and robustness to the disappearance and loss of the target as such will\ninevitably happen. For motivation, we have taken 10 randomly selected\nOTB-sequences, doubled each by attaching a reversed version and repeated each\ndouble sequence 20 times. On most of these repetitive videos, the best current\ntracker performs worse on each loop. This illustrates the difference between\noptimization for short-term versus long-term tracking. In a long-term tracker a\ncombined global and local search strategy is beneficial, allowing for recovery\nfrom failures and disappearance. Most importantly, the proposed tracker also\nemploys cautious updating, guided by self-quality assessment. The proposed\ntracker is still among the best on the 20-sec OTB-videos while achieving\nstate-of-the-art on the 100-sec UAV20L benchmark. On 10 new half-an-hour videos\nwith city bicycling, sport games etc, the proposed tracker outperforms others\nby a large margin where the 2010 TLD tracker comes second.","url_abs":"http://arxiv.org/abs/1711.10217v1","url_pdf":"http://arxiv.org/pdf/1711.10217v1.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":"tracking-for-half-an-hour","repo_url":"https://github.com/QUVA-Lab/Long-term-Siamese-Tracker","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}