Katsuhiko Ishiguro

Papers

Total Citations

H-Index

About

Katsuhiko Ishiguro is a leading researcher in robot audition and computational auditory scene analysis, with a focus on developing robust, integrated auditory functions for human-robot interaction. His work challenges traditional cascaded frameworks—where sound source localization, separation, and recognition are processed sequentially—by proposing unified, probabilistic models that improve overall system resilience. In his influential 2012 paper, "Unified auditory functions based on Bayesian topic model" (7 citations), Ishiguro introduced a novel approach that leverages Bayesian topic modeling to jointly handle multiple auditory tasks, reducing the need for environment-dependent tuning and mitigating performance degradation from subsystem failures. This contribution has been foundational for advancing robot audition in noisy, real-world settings. Beyond this work, Ishiguro has explored speech enhancement and active audition, contributing to systems that allow robots to adaptively focus on relevant sounds. His research bridges machine learning and robotics, offering elegant solutions to the challenge of making machines hear as flexibly as humans. For students and researchers, Ishiguro’s work exemplifies how probabilistic reasoning can transform fragmented auditory pipelines into cohesive, intelligent systems.