Satoshi Ukai

Nara Institute of Science and Technology

Papers

Total Citations

H-Index

About

Satoshi Ukai is a leading researcher in robot audition, with a primary focus on developing advanced signal processing techniques that enable robots to hear and interpret sound in real-world environments. His key contributions center on blind source separation (BSS) and sound scene decomposition for humanoid robots. Ukai pioneered a two-stage BSS method that combines SIMO-model-based independent component analysis (ICA) with binary masking, allowing robots to isolate individual sound sources from binaural mixed signals—a critical capability for interactive robots. His most cited work (32 citations) introduced this real-time system, demonstrating how robots can effectively separate speech and noise in dynamic settings. Ukai further advanced the field with his blind sound scene decomposition algorithm using SIMO-ICA, which enables robots to decompose complex auditory scenes without prior knowledge of the sound sources. These contributions have laid the groundwork for more sophisticated robot audition systems, directly impacting the development of socially interactive robots that can respond to human speech in noisy environments. His work remains foundational for researchers exploring auditory perception in robotics.