1 paper
Su Zhang, Ziyuan Zhao, Cuntai Guan
We used two multimodal models for continuous valence-arousal recognition using visual, audio, and linguistic information. The first model is the same as we used in ABAW2 and ABAW3,…