Title :
Hearing Is Believing: Biologically Inspired Methods for Robust Automatic Speech Recognition
Author :
Stern, Richard M. ; Morgan, Nelson
Author_Institution :
Dept. of Electr. & Comput. Eng., Carnegie Mellon Univ., Pittsburgh, PA, USA
Abstract :
The feature extraction stage of speech recognition is important historically and is the subject of much current research, particularly to promote robustness to acoustic disturbances such as additive noise and reverberation. Biologically inspired and biologically related approaches are an important subset of feature extraction methods for ASR.
Keywords :
feature extraction; speech recognition; acoustic disturbances; additive noise; automatic speech recognition; biologically inspired approach; biologically inspired methods; biologically related approach; feature extraction; reverberation; Adaptation models; Auditory systems; Automatic speech recognition; Computational modeling; Feature extraction; Gaussian processes; Physiology; Speech recognition;
Journal_Title :
Signal Processing Magazine, IEEE
DOI :
10.1109/MSP.2012.2207989
