2025
Multimodal Emotion Recognition with Target Speaker-Based Facial Embeddings
ICASSP 2025accepted
Effectively recognizing emotions requires sophisticated approaches for interpreting diverse modalities, particularly in real-world scenarios where multiple data sources, such as speech, text, and visual cues, are often noisy and incomplete. This study proposes an advanced multimodal emotion recognit…