{"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/rgb-t-object-trackingbenchmark-and-baseline","title":"RGB-T Object Tracking:Benchmark and Baseline","arxiv_id":"1805.08982","date":"2018-05-23","proceeding":null,"authors":["Chenglong Li","Xinyan Liang","Yijuan Lu","Nan Zhao","Jin Tang"],"abstract":"RGB-Thermal (RGB-T) object tracking receives more and more attention due to\nthe strongly complementary benefits of thermal information to visible data.\nHowever, RGB-T research is limited by lacking a comprehensive evaluation\nplatform. In this paper, we propose a large-scale video benchmark dataset for\nRGB-T tracking.It has three major advantages over existing ones: 1) Its size is\nsufficiently large for large-scale performance evaluation (total frame number:\n234K, maximum frame per sequence: 8K). 2) The alignment between RGB-T sequence\npairs is highly accurate, which does not need pre- or post-processing. 3) The\nocclusion levels are annotated for occlusion-sensitive performance analysis of\ndifferent tracking algorithms.Moreover, we propose a novel graph-based approach\nto learn a robust object representation for RGB-T tracking. In particular, the\ntracked object is represented with a graph with image patches as nodes. This\ngraph including graph structure, node weights and edge weights is dynamically\nlearned in a unified ADMM (alternating direction method of multipliers)-based\noptimization framework, in which the modality weights are also incorporated for\nadaptive fusion of multiple source data.Extensive experiments on the\nlarge-scale dataset are executed to demonstrate the effectiveness of the\nproposed tracker against other state-of-the-art tracking methods. We also\nprovide new insights and potential research directions to the field of RGB-T\nobject tracking.","url_abs":"http://arxiv.org/abs/1805.08982v1","url_pdf":"http://arxiv.org/pdf/1805.08982v1.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":null,"task_name":"8k"},{"task_slug":"object","task_name":"Object"},{"task_slug":"object-tracking","task_name":"Object Tracking"},{"task_slug":"rgb-t-tracking","task_name":"Rgb-T Tracking"}],"methods":[{"method_slug":"admm","method_name":"ADMM"}],"datasets_introduced":[{"slug":"rgbt234","name":"RGBT234","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1805.08982","atlas_url":"https://app.syntology.ai/?focus=1805.08982","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}