Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/100818
Title: Cross-lingual phone mapping for large vocabulary speech recognition of under-resourced languages
Authors: Do, Van Hai
Xiao, Xiong
Chng, Eng Siong
Li, Haizhou
Keywords: DRNTU::Engineering::Computer science and engineering
Issue Date: 2014
Source: DO, V. H., XIAO, X., CHNG, E. S., & LI, H. (2014). Cross-Lingual Phone Mapping for Large Vocabulary Speech Recognition of Under-Resourced Languages. IEICE Transactions on Information and Systems, E97.D(2), 285-295.
Series/Report no.: IEICE transactions on information and systems
Abstract: This paper presents a novel acoustic modeling technique of large vocabulary automatic speech recognition for under-resourced languages by leveraging well-trained acoustic models of other languages (called source languages). The idea is to use source language acoustic model to score the acoustic features of the target language, and then map these scores to the posteriors of the target phones using a classifier. The target phone posteriors are then used for decoding in the usual way of hybrid acoustic modeling. The motivation of such a strategy is that human languages usually share similar phone sets and hence it may be easier to predict the target phone posteriors from the scores generated by source language acoustic models than to train from scratch an under-resourced language acoustic model. The proposed method is evaluated using on the Aurora-4 task with less than 1 hour of training data. Two types of source language acoustic models are considered, i.e. hybrid HMM/MLP and conventional HMM/GMM models. In addition, we also use triphone tied states in the mapping. Our experimental results show that by leveraging well trained Malay and Hungarian acoustic models, we achieved 9.0% word error rate (WER) given 55 minutes of English training data. This is close to the WER of 7.9% obtained by using the full 15 hours of training data and much better than the WER of 14.4% obtained by conventional acoustic modeling techniques with the same 55 minutes of training data.
URI: https://hdl.handle.net/10356/100818
http://hdl.handle.net/10220/19586
ISSN: 0916-8532
DOI: 10.1587/transinf.E97.D.285
Schools: School of Computer Engineering 
Research Centres: Temasek Laboratories 
Rights: © 2014 The Institute of Electronics, Information and Communication Engineers. This paper was published in IEICE Transactions on Information and Systems and is made available as an electronic reprint (preprint) with permission of The Institute of Electronics, Information and Communication Engineers. The paper can be found at the following official DOI: http://dx.doi.org/10.1587/transinf.E97.D.285.  One print or electronic copy may be made for personal use only. Systematic or multiple reproduction, distribution to multiple locations via electronic or other means, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper is prohibited and is subject to penalties under law.
Fulltext Permission: open
Fulltext Availability: With Fulltext
Appears in Collections:SCSE Journal Articles
TL Journal Articles

SCOPUSTM   
Citations 20

13
Updated on Mar 2, 2024

Web of ScienceTM
Citations 20

10
Updated on Oct 27, 2023

Page view(s) 10

821
Updated on Mar 18, 2024

Download(s) 10

352
Updated on Mar 18, 2024

Google ScholarTM

Check

Altmetric


Plumx

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.