Understanding Identity Continuity in Thermal Video through Scene-Level Consistency 事件

Understanding Identity Continuity in Thermal Video through Scene-Level Consistency arXiv:2606.01694v1 Announce Type: new Abstract: Thermal pedestrian MOT remains challenging because weak appearance cues and frequent detection interruptions cause severe trajectory fragmentation. We study whether lightweight post-processing can recover identity continuity without relying on heavy re-identification models or complex online association. Starting from a YOLOv8 and SORT baseline, we add a modular ide