Takuya Narihira

Papers

Total Citations

H-Index

About

Takuya Narihira is a leading researcher in robotic motion planning and path planning, with a focus on integrating deep learning to overcome long-standing computational bottlenecks. His work centers on developing novel frameworks that combine sampling-based planners with deep neural networks, particularly 3D convolutional neural networks (3D-CNNs), to generate heuristic guidance for faster, more reliable collision-free pathfinding. His 2020 paper, "3D-CNN Based Heuristic Guided Task-Space Planner for Faster Motion Planning," with 26 citations, introduced a pioneering approach that significantly reduces planning time in complex, high-dimensional environments. Building on this, his 2019 work, "Fully Convolutional Search Heuristic Learning for Rapid Path Planners" (11 citations), advanced the field by learning search heuristics end-to-end, enabling real-time performance in large spaces with local traps. Narihira’s contributions are critical for applications ranging from mobile robot navigation to robotic arm manipulation, where speed and reliability are paramount. His research has been widely recognized for bridging the gap between classical planning algorithms and modern deep learning, offering practical solutions for autonomous systems operating under real-world constraints.