Improving Speech Emotion Recognition with Mutual Information Regularized Generative Model
This research introduces a novel framework leveraging mutual information regularization and cross-modal transfer to significantly boost speech emotion recogn...