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...

Level: advanced

By Anh-Tuan Mai, Cam-Van Thi Nguyen, Duc-Trong Le

Category: research