In this letter, a novel perceptually motivated single channel speech enhancement approach based on Deep Neural Network (DNN) is presented. Taking into account the good masking properties of the human auditory system, a new DNN architecture is proposed to reduce the perceptual effect of the residual noise. This new DNN architecture is directly trained to learn a gain function which is used to estimate the power spectrum of clean speech and shape the spectrum of the residual noise at the same time. Experimental results demonstrate that the proposed perceptually motivated speech enhancement approach could achieve better objective speech quality when tested with TIMIT sentences corrupted by various types of noise, no matter whether the noise conditions are included in the training set or not.
Wei HAN
PLA University of Science and Technology
Xiongwei ZHANG
PLA University of Science and Technology
Gang MIN
PLA University of Science and Technology
Meng SUN
PLA University of Science and Technology
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Wei HAN, Xiongwei ZHANG, Gang MIN, Meng SUN, "A Perceptually Motivated Approach for Speech Enhancement Based on Deep Neural Network" in IEICE TRANSACTIONS on Fundamentals,
vol. E99-A, no. 4, pp. 835-838, April 2016, doi: 10.1587/transfun.E99.A.835.
Abstract: In this letter, a novel perceptually motivated single channel speech enhancement approach based on Deep Neural Network (DNN) is presented. Taking into account the good masking properties of the human auditory system, a new DNN architecture is proposed to reduce the perceptual effect of the residual noise. This new DNN architecture is directly trained to learn a gain function which is used to estimate the power spectrum of clean speech and shape the spectrum of the residual noise at the same time. Experimental results demonstrate that the proposed perceptually motivated speech enhancement approach could achieve better objective speech quality when tested with TIMIT sentences corrupted by various types of noise, no matter whether the noise conditions are included in the training set or not.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.E99.A.835/_p
Copy
@ARTICLE{e99-a_4_835,
author={Wei HAN, Xiongwei ZHANG, Gang MIN, Meng SUN, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={A Perceptually Motivated Approach for Speech Enhancement Based on Deep Neural Network},
year={2016},
volume={E99-A},
number={4},
pages={835-838},
abstract={In this letter, a novel perceptually motivated single channel speech enhancement approach based on Deep Neural Network (DNN) is presented. Taking into account the good masking properties of the human auditory system, a new DNN architecture is proposed to reduce the perceptual effect of the residual noise. This new DNN architecture is directly trained to learn a gain function which is used to estimate the power spectrum of clean speech and shape the spectrum of the residual noise at the same time. Experimental results demonstrate that the proposed perceptually motivated speech enhancement approach could achieve better objective speech quality when tested with TIMIT sentences corrupted by various types of noise, no matter whether the noise conditions are included in the training set or not.},
keywords={},
doi={10.1587/transfun.E99.A.835},
ISSN={1745-1337},
month={April},}
Copy
TY - JOUR
TI - A Perceptually Motivated Approach for Speech Enhancement Based on Deep Neural Network
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 835
EP - 838
AU - Wei HAN
AU - Xiongwei ZHANG
AU - Gang MIN
AU - Meng SUN
PY - 2016
DO - 10.1587/transfun.E99.A.835
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E99-A
IS - 4
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - April 2016
AB - In this letter, a novel perceptually motivated single channel speech enhancement approach based on Deep Neural Network (DNN) is presented. Taking into account the good masking properties of the human auditory system, a new DNN architecture is proposed to reduce the perceptual effect of the residual noise. This new DNN architecture is directly trained to learn a gain function which is used to estimate the power spectrum of clean speech and shape the spectrum of the residual noise at the same time. Experimental results demonstrate that the proposed perceptually motivated speech enhancement approach could achieve better objective speech quality when tested with TIMIT sentences corrupted by various types of noise, no matter whether the noise conditions are included in the training set or not.
ER -