Downloads: 272

Files in This Item:
File Description SizeFormat 
s13634-015-0246-6.pdf2.11 MBAdobe PDFView/Open
Title: Reverberant speech recognition combining deep neural networks and deep autoencoders augmented with a phone-class feature
Authors: Mimura, Masato
Sakai, Shinsuke
Kawahara, Tatsuya  kyouindb  KAKEN_id  orcid https://orcid.org/0000-0002-2686-2296 (unconfirmed)
Author's alias: 三村, 正人
Keywords: Reverberant speech recognition
Deep Neural Networks (DNN)
Deep Autoencoder (DAE)
Issue Date: 23-Jul-2015
Publisher: SpringerOpen
Journal title: EURASIP Journal on Advances in Signal Processing
Volume: 2015
Issue: 1
Thesis number: 62
Abstract: We propose an approach to reverberant speech recognition adopting deep learning in the front-end as well as b a c k-e n d o f a r e v e r b e r a n t s p e e c h r e c o g n i t i o n s y s t e m, a n d a n o v e l m e t h o d t o i m p r o v e t h e d e r e v e r b e r a t i o n p e r f o r m a n c e of the front-end network using phone-class information. At the front-end, we adopt a deep autoencoder (DAE) for enhancing the speech feature parameters, and speech recognition is performed in the back-end using DNN-HMM acoustic models trained on multi-condition data. The system was evaluated through the ASR task in the Reverb Challenge 2014. The DNN-HMM system trained on the multi-condition training set achieved a conspicuously higher word accuracy compared to the MLLR-adapted GMM-HMM system trained on the same data. Furthermore, feature enhancement with the deep autoencoder contributed to the improvement of recognition accuracy especially in the more adverse conditions. While the mapping between reverberant and clean speech in DAE-based dereverberation is conventionally conducted only with the acoustic information, we presume the mapping is also dependent on the phone information. Therefore, we propose a new scheme (pDAE), which augments a phone-class feature to the standard acoustic features as input. Two types of the phone-class feature are investigated. One is the hard recognition result of monophones, and the other is a soft representation derived from the posterior outputs of monophone DNN. The augmented feature in either type results in a significant improvement (7–8 % relative) from the standard DAE.
Rights: © 2015 Mimura et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
URI: http://hdl.handle.net/2433/201887
DOI(Published Version): 10.1186/s13634-015-0246-6
Appears in Collections:Journal Articles

Show full item record

Export to RefWorks


Export Format: 


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.