Jangsik Park

TIV: A Tri-Modal Text–Image–Video Architecture for Crowd Behavior Classification and Captioning

This work proposes TIV (Text–Image–Video), a fusion-centric tri-modal framework dedicated entirely to optimized joint encoding that orchestrates frozen heterogeneous experts by integrating temporal video data via VideoMAE alongside localized spatial cues via a CLIP image encoder, and a text-based semantic guide via a t...