Kosuke Tahara

National Institute of Technology, Toyota College

Papers

Total Citations

H-Index

About

Kosuke Tahara is a robotics researcher advancing the frontier of visual navigation for autonomous mobile systems. His primary research areas lie at the intersection of computer vision and robot control, with a particular focus on enabling robots to navigate complex, changing environments using only visual input. His most cited work, "Probabilistic Visual Navigation with Bidirectional Image Prediction" (2021, 8 citations), tackles the fundamental challenge of robustly following a visual trajectory—a sequence of images akin to a video—despite substantial environmental changes or obstacles. Tahara’s key contribution is developing probabilistic models that allow a robot equipped solely with a single RGB fisheye camera to predict and adapt its path bidirectionally, mimicking the human ability to navigate by sight alone. This work demonstrates his commitment to creating practical, sensor-minimal solutions for real-world deployment. While his citation count is still growing, Tahara’s research represents a significant step toward endowing mobile robots with the visual intelligence needed for tasks ranging from autonomous delivery to search-and-rescue operations. His innovative approach to combining image prediction with probabilistic reasoning marks him as an emerging voice in the field of vision-based robotics.