Improvement of mask-based speech source separation using DNN.

Ge Zhan Zhaoqiong Huang Dongwen Ying Jielin Pan Yonghong Yan

Published in: ISCSLP (2016)

Keyphrases

source separation
audio features
blind source separation
independent component analysis
speech signal
denoising
single channel
speech recognition
low level
non stationary
audio visual
temporal structure
speaker identification