Toward Robust Multimodal Learning using Multimodal Foundational Models

Xianbing Zhao, Soujanya Poria, Xuejiao Li +2

Existing multimodal sentiment analysis tasks are highly rely on the assumption that the training and test sets are complete multimodal data, while this assumption can be difficult…