This book presents cutting-edge research in cross-disciplinary artificial intelligence for modeling human behavior. Real-world interactions, such as autonomous driving, speech communication, human-computer interaction, and sports behavior, require the prediction, interpretation, and control of complex interactions between agents and their environments. Although deep learning has achieved remarkable progress, it still faces two major challenges in real-world behavior modeling: ensuring interpretability and reducing the need for exhaustive data collection.

To address these challenges, this book brings together data-driven methods, mathematical models, domain knowledge, and computational approaches for understanding the environment, individuality, movement, values, and their interrelationships in human behavior. Rather than treating each application area separately, the volume highlights common principles and methodological connections across different domains.

The book consists of five parts. Part I introduces behavior signal processing from a cross-disciplinary perspective. Parts II, III, IV, and V review recent advances in vehicle signal processing, audio, speech, and language processing, human-computer interaction, and sports behavior modeling, respectively. The book will be useful for researchers, graduate students, and professionals interested in artificial intelligence, machine learning, behavior modeling, real-world data analysis, and interpretable prediction and control.

Kazuya Takeda is Professor at the Graduate School of Informatics, Nagoya University. His research covers behavior signal processing, driving behavior modeling, speech and acoustic signal processing, human-machine interaction, and sports behavior analysis. He has led cross-disciplinary research projects connecting artificial intelligence, real-world data, mobility, and human behavior modeling.

Keisuke Fujii is Associate Professor at the Graduate School of Informatics, Nagoya University. His research focuses on machine learning, multi-agent systems, sports science, collective motion, computational biology, and robotics. He develops data-driven and domain-informed methods for analyzing, predicting, and controlling multi-body time-series data in sports, animals, and other real-world systems.

Yoshio Ishiguro is Associate Professor at the Interfaculty Initiative in Information Studies, The University of Tokyo. His research interests include human-computer interaction, human augmentation, mixed and augmented reality, and infotainment systems for autonomous vehicles. He has worked across academia and industry on interaction design for intelligent mobility and human-centered systems.

Kento Ohtani is Designated Assistant Professor at the Graduate School of Informatics, Nagoya University. His research focuses on acoustic signal processing, sound source separation, three-dimensional audio, and audio technologies for controlling auditory impressions and user experiences. His work contributes to the broader integration of sound, perception, and behavior modeling.