Moa Lee
Papers
2
Total Citations
7
H-Index
2
About
Moa Lee is a researcher at the forefront of human-robot interaction, specializing in robust automatic speech recognition (ASR) for intelligent, motor-driven robots. Her primary research addresses a critical challenge: how can a robot understand human speech when its own mechanical movements—from fans and motors—generate disruptive "ego-noise"? Lee’s major contribution lies in developing deep neural network (DNN) architectures that leverage the robot’s own motor state as auxiliary information to suppress this noise. In her most-cited work (2018), she proposed a DNN-based system that integrates background noise suppression with acoustic modeling, using the on/off state of a head-shaking robot’s motor to dynamically filter its own sounds. She further advanced this concept in 2020 by augmenting latent DNN features to improve ASR robustness against ego-noise from moving or shaking robots. Though her citation counts (3–4 per paper) reflect a specialized, emerging field, her work is foundational for enabling seamless voice control in interactive robots. By teaching machines to listen despite their own chatter, Lee is paving the way for more natural, responsive robotic companions.
Research Focus
Key Achievements
Top Papers
- 1
- 2