Fetching the paper…

A cross-modal fusion network based on self-attention and residual structure for multimodal emotion recognition · Around