{"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/exploit-the-connectivity-multi-object","title":"Exploit the Connectivity: Multi-Object Tracking with TrackletNet","arxiv_id":"1811.07258","date":"2018-11-18","proceeding":null,"authors":["Gaoang Wang","Yizhou Wang","Haotian Zhang","Renshu Gu","Jenq-Neng Hwang"],"abstract":"Multi-object tracking (MOT) is an important and practical task related to\nboth surveillance systems and moving camera applications, such as autonomous\ndriving and robotic vision. However, due to unreliable detection, occlusion and\nfast camera motion, tracked targets can be easily lost, which makes MOT very\nchallenging. Most recent works treat tracking as a re-identification (Re-ID)\ntask, but how to combine appearance and temporal features is still not well\naddressed. In this paper, we propose an innovative and effective tracking\nmethod called TrackletNet Tracker (TNT) that combines temporal and appearance\ninformation together as a unified framework. First, we define a graph model\nwhich treats each tracklet as a vertex. The tracklets are generated by\nappearance similarity with CNN features and intersection-over-union (IOU) with\nepipolar constraints to compensate camera movement between adjacent frames.\nThen, for every pair of two tracklets, the similarity is measured by our\ndesigned multi-scale TrackletNet. Afterwards, the tracklets are clustered into\ngroups which represent individual object IDs. Our proposed TNT has the ability\nto handle most of the challenges in MOT, and achieve promising results on MOT16\nand MOT17 benchmark datasets compared with other state-of-the-art methods.","url_abs":"http://arxiv.org/abs/1811.07258v1","url_pdf":"http://arxiv.org/pdf/1811.07258v1.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":"exploit-the-connectivity-multi-object","repo_url":"https://github.com/GaoangW/TNT","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"autonomous-driving","task_name":"Autonomous Driving"},{"task_slug":"multi-object-tracking","task_name":"Multi-Object Tracking"},{"task_slug":"object","task_name":"Object"},{"task_slug":"object-tracking","task_name":"Object Tracking"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/multi-object-tracking-on-mot16","task":"Multi-Object Tracking","dataset":"MOT16","model":"TNT","rank_in_archive_order":20,"of":24,"metrics":{"MOTA":"49.2"},"uses_additional_data":false},{"leaderboard":"/sota/multi-object-tracking-on-mot17","task":"Multi-Object Tracking","dataset":"MOT17","model":"TNT","rank_in_archive_order":46,"of":48,"metrics":{"MOTA":"51.9"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.07258","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}