Beyond Stepwise Modeling: Toward a Unified Contextual Reasoning Framework for Hyperspectral Video Object Tracking
Hyperspectral video data provide complementary contextual cues across spectral, spatial, and temporal dimensions for modeling object dynamics under challenging conditions. Many existing hyperspectral video object tracking (HVOT) approaches organize spatial-spectral and temporal modeling in successive stages, leaving ro...