Asynchronous Multimodal Video Sequence Fusion via Learning Modality-Exclusive and -Agnostic Representations

Dingkang Yang, Mingcheng Li, Linhao Qu +4

Understanding human intentions (e.g., emotions) from videos has received considerable attention recently. Video streams generally constitute a blend of temporal data stemming from…