Divide and Refine: Enhancing Multimodal Representation and Explainability for Emotion Recognition in Conversation
Explore the Divide and Refine framework, a novel architecture that decomposes speech, facial, and textual data to enhance emotion recognition accuracy and in...