A Survey:Emotion Recognition in the Intelligent Cockpit

As intelligent cockpits evolve toward human-vehicle mutual adaptation, recognizing the physiological and emotional states of drivers and passengers has become essential for enhancing safety and user experience. This paper reviews the role, methods, and advancements of emotion recognition in intelligent cockpit scenarios, as well as the cultural differences between Eastern and Western contexts. Applications in driver monitoring, passenger experience, and human-machine interaction are examined, highlighting benefits such as proactive alerts, personalized services, and affective design. The review covers four key modalities: facial expressions, speech, physiological signals, and driving behavior. Recent trends include hybrid models for facial analysis, robust speech recognition under low expressiveness, multimodal physiological sensing with federated learning, and behavior-based emotion inference. Key challenges such as multimodal fusion, cross-scenario generalization, and edge deployment are discussed. Future directions involve lightweight, personalized systems enabled by federated learning, flexible sensors, and attention mechanisms.