Tokyo University of Agriculture and Technology Tanaka & Kataoka Lab.
Social Behavior Analysis Using Deep Learning
Entomophagy, as one form of dietary diversity, has been traditionally practiced in many regions worldwide, with over 2 billion people reportedly consuming insects regularly today. In this research, we employ deep learning to elucidate the molecular foundations of social behavior, using insects as model organisms. By analyzing behavior through AI technology from multiple angles, we comprehensively understand subtle behavioral patterns and inter-individual interactions that have been historically difficult to observe (Fig. 4,5).
Specifically, we combine multiple deep learning technologies to improve behavioral tracking accuracy and establish a foundation for understanding the molecular mechanisms of specific social behaviors (e.g., courtship behavior, fighting behavior). Furthermore, by integrating behavioral data with omics data, we systematize the molecular basis of social behavior into a predictive model that reveals how individual and environmental factors influence social behavior.
This research proposes novel methodologies through the integration of behavioral biology and information science. It is expected to deepen our understanding of the evolution and adaptation of animal behavior while contributing to the development of next-generation behavioral control technologies.
