This book presents a comprehensive theoretical and methodological framework for cross-scenario cognitive computing in intelligent driving systems. Moving beyond traditional perception-centered autonomous driving approaches, the book explores how cognitive intelligence can be integrated into intelligent transportation systems through multimodal sensing, physiological signal processing, machine learning, and behavioral understanding. It systematically addresses driver emotion recognition, cognitive workload analysis, behavioral intention detection, vigilance monitoring, and corner-case perception in autonomous driving environments. The book combines theoretical foundations with practical methodologies and experimental validation. It introduces advanced techniques including frequency-aware temporal modeling, hypergraph convolutional networks, manifold learning, multimodal large language models, and latent-flow diversification methods for safety-critical scenario generation. Designed for researchers, graduate students, and industry practitioners, the book serves as a valuable reference for intelligent driving, artificial intelligence, biomedical engineering, and cognitive computing applications in future transportation systems.
