1
Return

STCTracker: Enhancing sequential temporal consistency in referring multi-object tracking

delete2026-06-10
delete0
PRE
AI
H
Hong Zhang
J
Jiabi Zhao
D
Ding Yuan
H
Hanyang Liu
Y
Yang Han
Y
Y. F. Yang *
DOI：10.1016/j.imavis.2026.106073delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• DynaSORT, an enhanced multi-object tracker, integrates a Movement Vector Constraint, Bifactor Kalman Filter, and Occlusion-aware Exponential Moving Average to improve temporal ID consistency. • DRM, an enhanced referring module, employs dual auxiliary discriminators to promote temporal referring consistency. • STCTracker, a referring multi-object tracking method, integrates DynaSORT and DRM in a modular framework. • The proposed method achieves state-of-the-art performance on the Refer-KITTI and KITTI benchmarks.

Journal

Image and Vision Computing cover
Image and Vision Computing
IF:
4.2
Papers:
4.0K
Citations:
6.7K

Organization

No organization information available
Cited Papers

Cited Papers

Citing Papers

Citing Papers