Chokushi Yuuto

Papers

Total Citations

H-Index

About

Chokushi Yuuto is a computer vision and robotics researcher whose work centers on visual place recognition, long-term visual SLAM (Simultaneous Localization and Mapping), and image-based scene understanding. His research addresses one of the most persistent challenges in autonomous navigation: enabling robots to reliably recognize locations across dramatically different environmental conditions, such as seasonal changes in appearance. His most influential contribution, "Leveraging image-based prior in cross-season place recognition" (2015, 21 citations), introduced a compact discriminative scene descriptor capable of handling significant appearance variation — a critical capability for real-world robotic deployment. Building on this, his 2014 work on "Mining visual phrases for long-term visual SLAM" (13 citations) advanced scene description beyond conventional bag-of-words approaches, demonstrating improved robustness in dynamic and partially changing environments. His earlier research on object-level image retrieval through a Bag-of-Bounding-Boxes framework further reflects his interest in semantically meaningful, efficient scene representations. Across these works, Yuuto has consistently pursued compact yet discriminative descriptors that bridge perception and long-term autonomy, contributing meaningfully to the foundations of robust robot localization in real-world, ever-changing environments.