Enhancing Multimodal Emotion Recognition via Multi-feature Encoding and Attention-Based Fusion
A novel multimodal emotion recognition framework that integrates rich audio and visual feature extraction with an attention-based fusion strategy, and introduces a feature-level fusion mechanism based on multi-head attention to enhance multimodal synergy.
Xu Lin, Ke Wang, Huibin Kang et al.
· International Conference on... · 1 citation