.png){kind=icon}
.png){kind=icon}
Return
CausTrack: Visual tracking via structural causality and deconfounded feature learning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2025.132340.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We propose CausTrack, a causality-based Transformer for robust and interpretable visual tracking. • We design three modules for causal temporal modeling, deconfounded features, and reliable estimation. • CausTrack achieves state-of-the-art performance on seven benchmarks at 28 FPS on a V100 GPU.
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W
Organization
No organization information available