An auditory feature extraction method based on forward-masking and its application in robust speaker identification and speech recognition

摘要

This article presents a new auditory feaure extraction method, which consideres the forward-masking mechanism of auditory nerves and feasible in practice. Two features based on this method are extracted: FMFRC (forward masking firing-rate cepstrum) and FMSRC (forward masking synchronized rate cepstrum). Isolate-word speech recognition and text-dependent speaker identification experiments based on TI46 are conducted. The experiment results show that the new auditory features has comparable performance with MFCC under clean environment but far better noise-resistant property than MFCC in both tasks.