Max Robotics — Global Service Robot Directory + US Certification

About

Masatoshi Ando is a computer vision and robotics researcher whose work centers on visual place recognition, long-term simultaneous localization and mapping (SLAM), and scene understanding for autonomous systems. His research addresses one of the most persistent challenges in robot navigation: enabling machines to reliably recognize locations across dramatically changing conditions, such as seasonal variation and shifting environmental appearances. Ando's most influential contribution, "Leveraging image-based prior in cross-season place recognition" (2015, 21 citations), introduced a compact discriminative scene descriptor that leverages object-level priors to maintain recognition accuracy despite significant visual change. This work builds on his earlier research into "visual phrases" for long-term visual SLAM (2014, 13 citations), where he proposed an alternative to conventional bag-of-words descriptors — a more semantically grounded approach capable of handling semi-dynamic and partially changing environments. His 2013 work on bag-of-bounding-boxes further advanced object-level image retrieval for robotic applications, offering a semantic and computationally efficient framework for scene matching. Collectively, Ando's contributions have shaped how autonomous systems perceive and remember places over time, making him a notable contributor to the intersection of computer vision, semantic scene representation, and practical robot localization.