{"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/a-novel-performance-evaluation-methodology","title":"A Novel Performance Evaluation Methodology for Single-Target Trackers","arxiv_id":"1503.01313","date":"2015-03-04","proceeding":null,"authors":["Matej Kristan","Jiri Matas","Ales Leonardis","Tomas Vojir","Roman Pflugfelder","Gustavo Fernandez","Georg Nebehay","Fatih Porikli","Luka Cehovin"],"abstract":"This paper addresses the problem of single-target tracker performance\nevaluation. We consider the performance measures, the dataset and the\nevaluation system to be the most important components of tracker evaluation and\npropose requirements for each of them. The requirements are the basis of a new\nevaluation methodology that aims at a simple and easily interpretable tracker\ncomparison. The ranking-based methodology addresses tracker equivalence in\nterms of statistical significance and practical differences. A fully-annotated\ndataset with per-frame annotations with several visual attributes is\nintroduced. The diversity of its visual properties is maximized in a novel way\nby clustering a large number of videos according to their visual attributes.\nThis makes it the most sophistically constructed and annotated dataset to date.\nA multi-platform evaluation system allowing easy integration of third-party\ntrackers is presented as well. The proposed evaluation methodology was tested\non the VOT2014 challenge on the new dataset and 38 trackers, making it the\nlargest benchmark to date. Most of the tested trackers are indeed\nstate-of-the-art since they outperform the standard baselines, resulting in a\nhighly-challenging benchmark. An exhaustive analysis of the dataset from the\nperspective of tracking difficulty is carried out. To facilitate tracker\ncomparison a new performance visualization technique is proposed.","url_abs":"http://arxiv.org/abs/1503.01313v3","url_pdf":"http://arxiv.org/pdf/1503.01313v3.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":[],"tasks":[{"task_slug":"clustering","task_name":"Clustering"}],"methods":[],"datasets_introduced":[{"slug":"vot2014","name":"VOT2014","full_name":"Visual Object Tracking Challenge 2014"}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1503.01313","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}