Summary: MULTIPLE­MICROPHONE TIME­VARYING FILTERS FOR ROBUST SPEECH
RECOGNITION
Calvin Yiu­Kit Lai Parham Aarabi
The Artificial Perception Lab
The Edward S. Rogers Sr. Department of Electrical and Computer Engineering
University of Toronto
calvin.lai@utoronto.ca parham@ecf.utoronto.ca
ABSTRACT
A multiple microphone time varying filter that is an ex­
tension of the dual­microphone speech enhancement tech­
nique of [7] is proposed and experimentally analyzed. The
technique utilizes information regarding the locations of the
speech source of interest and the microphones to compute a
time varying filter that results in substantial noise reduction
over other speech enhancement techniques such as delay­
and­sum beamformingand superdirective beamforming. For
example, digit recognition results in an environment with
two speakers and a reverberation time of 0.1s show a recog­
nition accuracy rate increase of 25.2% over delay­and­sum
beamforming and an increase of 26.5% over superdirective

Source: Aarabi, Parham - Department of Electrical and Computer Engineering, University of Toronto

Collections: Computer Technologies and Information Sciences