Title:Multi-modal Multi-label Facial Action Unit Detection with Transformer

Abstract:Facial Action Coding System is an important approach of facial expression this http URL paper describes our submission to the third Affective Behavior Analysis (ABAW) 2022 competition. We proposed a transfomer based model to detect facial action unit (FAU) in video. To be specific, we firstly trained a multi-modal model to extract both audio and visual feature. After that, we proposed a action units correlation module to learn relationships between each action unit labels and refine action unit detection result. Experimental results on validation dataset shows that our method achieves better performance than baseline model, which verifies that the effectiveness of proposed network.