Yasunari Obuchi

Papers

Total Citations

H-Index

About

Yasunari Obuchi is a leading researcher in speech processing for human-robot interaction, with a focus on enabling robots to understand human speech in real-world, noisy environments. His work is foundational to the development of human-symbiotic robots—machines designed to live and work alongside people. Obuchi’s major contributions include pioneering robust automatic speech recognition (ASR) for robots, as demonstrated in his 2007 paper on the EMIEW robot (15 citations), which tackled the critical challenge of degraded accuracy in noisy settings. He also advanced direction-of-arrival (DOA) estimation with innovative methods like the "modified delay-and-sum beamformer based on sparseness" (MDSBF) and "stepwise phase difference restoration" (SPIRE), detailed in his 2009 work (11 citations), enabling robots to locate speakers in three dimensions. More recently, his 2022 paper (2 citations) introduced a novel web-questionnaire-based method for creating spontaneous speech corpora, assuming humans as speech targets, which is vital for training natural language models. Obuchi’s research directly impacts the future of assistive and service robotics, bridging the gap between laboratory conditions and dynamic human environments.