Track to Detect and Segment: An Online Multi-Object Tracker

被引：238
作者：
Wu, Jialian [1 ]
Cao, Jiale [2 ]
Song, Liangchen [1 ]
Wang, Yu [3 ]
Yang, Ming [3 ]
Yuan, Junsong [1 ]
机构：
[1] SUNY Buffalo, Buffalo, NY 14260 USA
[2] TJU, Tianjin, Peoples R China
[3] Horizon Robot, Beijing, Peoples R China
来源：
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021 | 2021年
基金：
美国国家科学基金会;
关键词：
D O I：
10.1109/CVPR46437.2021.01217
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
Most online multi-object trackers perform object detection stand-alone in a neural net without any input from tracking. In this paper, we present a new online joint detection and tracking model, TraDeS (TRAck to DEtect and Segment), exploiting tracking clues to assist detection end-to-end. TraDeS infers object tracking offset by a cost volume, which is used to propagate previous object features for improving current object detection and segmentation. Effectiveness and superiority of TraDeS are shown on 4 datasets, including MOT (2D tracking), nuScenes (3D tracking), MOTS and Youtube-VIS (instance segmentation tracking).
引用
收藏
页码：12347 / 12356
页数：10
相关论文
共 66 条
[1]  
[Anonymous], 2018, CVPR
[2]  
[Anonymous], 2019, CVPR, DOI DOI 10.1109/CVPR.2019.01298
[3]  
[Anonymous], 2017, CVPR, DOI DOI 10.1109/CVPR.2017.394
[4]  
[Anonymous], 2019, CVPR, DOI DOI 10.1109/CVPR.2019.00971
[5]  
[Anonymous], 2020, IROS, DOI DOI 10.1109/IROS45743.2020.9341164
[6]  
[Anonymous], 2017, CVPR, DOI DOI 10.1109/CVPR.2017.615
[7]  
[Anonymous], 2020, CVPR, DOI DOI 10.1109/CVPR42600.2020.00653
[8]  
[Anonymous], 2020, CVPR, DOI DOI 10.1109/CVPR42600.2020.00628
[9]  
[Anonymous], 2016, CVPR, DOI DOI 10.1109/CVPR.2016.509
[10]  
[Anonymous], 2018, CVPR, DOI DOI 10.1109/CVPR.2018.00255
1234567