Projects per year
Abstract
State-of-the-art multi-object tracking (MOT) methods follow the tracking-by-detection paradigm, where object trajectories are obtained by associating per-frame outputs of object detectors. In crowded scenes, however, detectors often fail to obtain accurate detections due to heavy occlusions and high crowd density. In this paper, we propose a new MOT paradigm, tracking-by-counting, tailored for crowded scenes. Using crowd density maps, we jointly model detection, counting, and tracking of multiple targets as a network flow program, which simultaneously finds the global optimal detections and trajectories of multiple targets over the whole video. This is in contrast to prior MOT methods that either ignore the crowd density and thus are prone to errors in crowded scenes, or rely on a suboptimal two-step process using heuristic density-aware point-tracks for matching targets. Our approach yields promising results on public benchmarks of various domains including people tracking, cell tracking, and fish tracking.
| Original language | English |
|---|---|
| Article number | 9298464 |
| Pages (from-to) | 1439-1452 |
| Journal | IEEE Transactions on Image Processing |
| Volume | 30 |
| Online published | 17 Dec 2020 |
| DOIs | |
| Publication status | Published - 2021 |
Research Keywords
- crowd density map
- flow tracking
- multiple people tracking
- People tracking
Publisher's Copyright Statement
- COPYRIGHT TERMS OF DEPOSITED POSTPRINT FILE: © 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Ren, W., Wang, X., Tian, J., Tang, Y., & Chan, A. B. (2021). Tracking-by-Counting: Using Network Flows on Crowd Density Maps for Tracking Multiple Targets. IEEE Transactions on Image Processing, 30, 1439-1452. Article 9298464. https://doi.org/10.1109/TIP.2020.3044219
Fingerprint
Dive into the research topics of 'Tracking-by-Counting: Using Network Flows on Crowd Density Maps for Tracking Multiple Targets'. Together they form a unique fingerprint.Projects
- 2 Finished
-
GRF: Wide-area Crowd Counting on Camera Networks using Multi-view Fusion
CHAN, A. B. (Principal Investigator / Project Coordinator)
1/09/18 → 27/02/23
Project: Research
-
TBRS: Safety, Reliability, and Disruption Management of High Speed Rail and Metro Systems
XIE, M. (Principal Investigator / Project Coordinator), BENSOUSSAN, A. (Co-Principal Investigator), LO, S. M. (Co-Principal Investigator), SHOU, B. (Co-Principal Investigator), SINGPURWALLA, N. D. (Co-Principal Investigator), TSE, W. T. P. (Co-Principal Investigator), TSUI, K. L. (Co-Principal Investigator), YU, Y. (Co-Principal Investigator), YUEN, K. K. R. (Co-Principal Investigator), CHAN, A. B. (Co-Investigator), CHAN, N.-H. (Co-Investigator), CHIN, K. S. (Co-Investigator), CHOW, H. A. (Co-Investigator), Chow, W. K. (Co-Investigator), EDESESS, M. (Co-Investigator), GOLDSMAN, D. M. (Co-Investigator), Huang, J. (Co-Investigator), LEE, W. M. (Co-Investigator), LI, L. (Co-Investigator), LI, C. L. (Co-Investigator), LING, M. H. A. (Co-Investigator), LIU, S. (Co-Investigator), MURAKAMI, J. (Co-Investigator), NG, S. Y. S. (Co-Investigator), NI, M. C. (Co-Investigator), TAN, M.H.-Y. (Co-Investigator), Wang, W. (Co-Investigator), Wang, J. (Co-Investigator), WONG, C. K. (Co-Investigator), WONG, S. Y. Z. (Co-Investigator), WONG, S. C. (Co-Investigator), Xu, Z. (Co-Investigator), ZHANG, Z. (Co-Investigator), Zhang, D. (Co-Investigator), ZHAO, J. L. (Co-Investigator) & Zhou, Q. (Co-Investigator)
1/01/16 → 31/12/21
Project: Research