AVTrack: Audio-Visual Tracking in Human-centric Complex Scenes [ICML 2026]
FudanCVL/AVTrack has added +9 stars since the first tracked point, with current momentum at 16.80.