Kousuke Yamaguchi

Papers

Total Citations

H-Index

About

Kousuke Yamaguchi is a computer vision researcher whose work focuses on advancing object-level scene understanding for autonomous systems. His most cited paper, "Leveraging Object Proposals for Object-Level Change Detection" (2018, 7 citations), introduces a novel approach that shifts change detection from traditional pixel- or keypoint-based methods to object-level analysis. By leveraging object proposals, Yamaguchi’s method achieves the speed required for real-time applications like self-driving cars and robotics, while providing more semantically meaningful change detection. This contribution addresses a critical gap in autonomous perception—enabling systems to understand not just that a scene has changed, but what specific objects have appeared, disappeared, or moved. Though his citation count is modest, his work is notable for its practical orientation toward deployment in dynamic environments. Yamaguchi’s research sits at the intersection of object detection, scene understanding, and efficient feature-based differencing, offering a bridge between low-level image processing and high-level semantic reasoning. His approach holds promise for improving the reliability of autonomous navigation and robotic manipulation in changing environments.