A NEW COST FUNCTION FOR DNN-BASED SPEECH ENHANCEMENT COMBINING NMF AND CASA

In this paper, a novel deep neural network (DNN) training approach is proposed for speech enhancement based on nonnegative matrix factorization (NMF) and computational auditory scene analysis (CASA). Considering a higher correlation of NMF algorithm along the frequency bins for the time-varying sign...

Full description

Saved in:
Bibliographic Details
Main Authors: Yan, Bofang, Bao, Changchun, Bai, Zhigang
Format: Conference Proceeding
Language:English
Subjects:
computational auditory scene analysis
Cost function
Deep neural network
Filter banks
Noise measurement
nonnegative matrix factorization
Speech enhancement
Time-frequency analysis
Training
Wiener filters
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, a novel deep neural network (DNN) training approach is proposed for speech enhancement based on nonnegative matrix factorization (NMF) and computational auditory scene analysis (CASA). Considering a higher correlation of NMF algorithm along the frequency bins for the time-varying signals and a high noise making effect of CASA, we propose a new cost function for DNN training, which consists of the ideal ratio mask (IRM) and NMF based Wiener-like filter. Extensive experiments are carried out to verify the performance of the proposed method. Moreover, we compare the performance of the developed algorithm with traditional NMF approach, NMF-based linear minimum mean square error (LMMSE) filter approach and CASA method. Our results demonstrate that the proposed approach improved speech quality greatly.
ISSN:2164-5221
DOI:10.1109/ICSP.2018.8652348