Evaluation and analysis of auditory model front ends for robust speech recognition program summary
Richard P. Lippmann · 1992
The purpose of this work is to integrate a number of auditory model front ends into a high-performance HMM recognizer, to test and evaluate these front ends on noisy speech, and to analyze the results in order to develop a more robust front end which may combine features of a number of the current auditory model-based systems.