Kohei Yatabe
Papers
2
Total Citations
18
H-Index
2
About
Kohei Yatabe is a researcher working at the intersection of audio-visual signal processing and machine learning, with a particular focus on sound source localization and self-supervised learning methodologies. His most recognized contribution lies in developing neural frameworks that enable autonomous systems to identify and locate sounding objects within visual scenes — a critical capability for robots navigating complex real-world environments. Recognizing the practical impossibility of exhaustively labeling the vast diversity of sounding objects encountered in everyday settings, Yatabe's work pioneered self-supervised approaches that leverage probabilistic spatial modeling, allowing systems to learn audio-visual correspondences without requiring manual annotation. This research addresses a fundamental bottleneck in scalable robot perception and has garnered notable attention within the robotics and audio-visual learning communities, accumulating citations that reflect its relevance to ongoing challenges in multimodal sensing. By combining principled probabilistic frameworks with modern neural architectures, Yatabe's contributions offer a compelling pathway toward more adaptable and autonomous perceptual systems, making his work of considerable interest to students and researchers in robotics, signal processing, and self-supervised machine learning alike.
Research Focus
Key Achievements
Top Papers
- 1
- 2