RTracker: Recoverable Tracking via PN Tree Structured Memory

arXiv:2403.19242v1 Announce Type: new
Abstract: Existing tracking methods mainly focus on learning better target representation or developing more robust prediction models to improve tracking performance. While tracking performance has significantly improved, the target loss issue occurs frequently due to tracking failures, complete occlusion, or out-of-view situations. However, considerably less attention is paid to the self-recovery issue of tracking methods, which is crucial for practical applications. To this end, we propose a recoverable tracking framework, RTracker, that uses a tree-structured memory …

abstract arxiv attention cs.cv focus however issue loss memory performance prediction prediction models representation robust tracking tree type via view