Fumihito Kyota

Papers

Total Citations

H-Index

About

Fumihito Kyota’s research lies at the intersection of autonomous robotics, humanoid agents, and intelligent grasping—a critical capability for machines that must interact with the physical world. His most cited work, “Detection and evaluation of grasping positions for autonomous agents” (2005, 37 citations), addresses a fundamental challenge: how can a humanoid agent, evolving autonomously in a virtual environment, determine the optimal portion of an object to grasp when given a simple command? Kyota’s contributions provide a systematic method for recognizing and evaluating graspable positions, enabling agents to make context-aware decisions without explicit human guidance. This work has influenced subsequent research in robotic manipulation and autonomous decision-making, particularly in simulated training environments. Beyond this paper, Kyota’s broader research explores how virtual agents can develop adaptive behaviors through environmental interaction, bridging computer graphics, artificial intelligence, and robotics. His approach emphasizes practical, real-time solutions for autonomous systems, making his findings valuable for developers of humanoid robots and interactive simulations. With a career focused on enabling machines to perceive and act upon their surroundings, Kyota continues to shape how autonomous agents learn to handle objects—a foundational skill for the next generation of intelligent robots.