Our Approach
As humans, we utilize various sensors, such as stereo microphones (ears) and stereo cameras (eyes), as well as an "intelligent" processor (brain), to learn and make decisions. Our natural intelligence, aided by our body sensors (for touch, smell, and feel), allows us to perform tasks such as:
- Determining direction and distance (source localization)
- Communicating in diverse and noisy environments (blind source separation - a cocktail party problem)
- Recognizing and identifying sounds (sound recognition)
- Understanding the overall environment based on various sounds (acoustic scene classification)
Ideally, we aim to equip "intelligent" humanoid robots with all of these capabilities. Therefore, we strive to contribute to the advancement of AI systems from the perspectives of signal processing and machine learning.