Beyond Visual Cues: Synchronously Exploring Target-Centric Semantics for Vision-Language Tracking

arXiv:2311.17085v2 Announce Type: replace
Abstract: Single object tracking aims to locate one specific target in video sequences, given its initial state. Classical trackers rely solely on visual cues, restricting their ability to handle challenges such as appearance variations, ambiguity, and distractions. Hence, Vision-Language (VL) tracking has emerged as a promising approach, incorporating language descriptions to directly provide high-level semantics and enhance tracking performance. However, current VL trackers have not fully exploited the power of VL learning, as they suffer from …

abstract arxiv beyond challenges cs.cv distractions language semantics state tracking type video vision visual visual cues