Comprehensible AI for Multimodal State Detection

Affective computing enables computers to recognize and respond to human emotions and cognitive states using multimodal state recognition. This chapter explores cognitive load estimation through a machine learning life cycle, addressing data collection and preparation, modeling, and deployment challenges. Experiments demonstrate effective unimodal and fusion models for predicting cognitive load. Applications in gaming, healthcare, and driver monitoring have been discussed, providing valuable insights for researchers in this field.