Multimodal sentiment analysis based on disentangled representation learning and cross-modal-context association mining

Zuhe Li, Panbo Liu, Yushan Pan, Weiping Ding, Jun Yu, Haoran Chen, Weihua Liu, Yiming Luo, Hao Wang

Abstract: Highlights•A multiview collaborative perception framework focusing on representation, mining, and cooperation.•A multimodal disentangled representation learning scheme for sentiment association mining.•A CMCAM module capturing cross-modal, intramodal, and intermodal sentiment associations.