Yui Sudo

Tokyo Institute of Technology, Honda (Japan)

Papers

Total Citations

H-Index

About

Yui Sudo is a researcher advancing the field of robot audition, focusing on how machines perceive and interpret complex acoustic environments. Her primary research areas include environmental sound segmentation, sound source localization, and acoustic transfer function modeling. Sudo’s major contribution is the development of Mask U-Net-based methods for environmental sound segmentation, which enable robots to isolate and identify overlapping sounds in real-world, noisy settings—a critical step toward robust human-robot interaction. Her 2020 paper on this topic has garnered 14 citations, while her 2019 foundational work has 11 citations, reflecting growing interest in her approach. More recently, Sudo has tackled the challenge of dynamic acoustic environments with her 2023 work on online adaptation of Fourier series-based acoustic transfer function models, which improves sound source localization and separation in real time. This innovation is particularly valuable for microphone array systems in robots, allowing them to adapt to changing soundscapes without recalibration. Sudo’s research bridges signal processing and artificial intelligence, offering practical solutions for robots to hear and respond as effectively as humans.