.png){kind=icon}
.png){kind=icon}
Return
LTSTrack: Visual tracking with long-term temporal sequence
.png){kind=icon}
PRE
AI
DOI：10.1016/j.patcog.2026.113052.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We propose a concise appearance compression for targets as tokens without extra module. • We introduce a temporal fusion module with Mamba blocks encoding long-term dependencies. • We present the LTSTrack that exploits long-term temporal context across 300 frames. • The proposed tracker achieves state-of-the-art performance on benchmarks with- out sacrificing computational efficiency.
Journal
IF:
7.6
Papers:
1.3W
Citations:
4.5W