Keisuke Fujii

Japan Science and Technology Agency, Nagoya University

Papers

Total Citations

H-Index

About

Keisuke Fujii is pioneering the intersection of artificial intelligence, sports analytics, and multi-agent systems, with a focus on decoding complex collective behaviors from limited data. His most impactful work introduces a data-driven framework for extracting and evaluating team tactics from football event and freeze-frame data, offering a novel lens to understand strategic spatial and action behaviors during possession. This research, already garnering 6 citations since 2024, bridges game theory and robotics by modeling how opposing multi-agent teams interact. Fujii has also advanced predictive world modeling, presenting the first open-vocabulary framework that learns causal mental simulations from sensor observations—a leap toward adaptable, human-like intelligence. Additionally, his work on learning navigational patterns from partial observations enables mobile robots to infer rule-constrained pathways, such as road lanes, from incomplete environmental data. By tackling challenges in partial observability and multi-agent coordination, Fujii’s contributions are shaping autonomous systems and sports analytics, with his 2024 papers already drawing attention for their foundational impact. His research not only advances AI but also offers practical tools for understanding dynamic, real-world team strategies.