ABSTRACT

Deep Neural Networks for Speech Separation with Application to Robust Speech Recognition   [open pdf - 1MB]

"This project will investigate the speech separation problem and apply the results of speech separation to robust automatic speech recognition (ASR). Speech separation has been recently formulated as a time-frequency masking problem, which shifts the research focus to supervised learning. The proposed effort will employ deep neural networks (DNN) as the learning machine for supervised separation The proposed research aims to achieve the following objectives. The first objective is separation of speech from background noise. This will be accomplished by training DNN classifiers on extracted acoustic-phonetic features. The second objective is integration of spectrotemporal context for improved separation performance. Conditional random fields will be used to encode contextual constraints. The third objective is to achieve robust ASR in the DNN framework through integrated acoustic modeling and separation. The performance of the proposed system will be systematically evaluated using the recently constructed CHIME-2corpus."

Report Number:
AFRL-RI-RS-TR-2018-152
Author:
Publisher:
Date:
2018-06
Copyright:
Public Domain
Retrieved From:
Defense Technical Information Center (DTIC): http://www.dtic.mil/dtic/
Format:
pdf
Media Type:
application/pdf
URL:
Help with citations