Kumar Ashutosh

video understanding (4) multimodal learning (3) zero-shot learning (2) action recognition (2) instructional video (2) contrastive learning (2) egocentric video (2) egocentric vision (2) self-supervised learning (1) pose estimation (1) hierarchical representation (1) audio-visual learning (1) video retrieval (1) human-object interaction (1) future prediction (1) question answering (1) hand pose estimation (1) regret minimization (1) 3d scene understanding (1) weakly supervised learning (1)