CASIA OpenIR  > 09年以前成果
Statistic model based dynamic channel compensation for telephony speech recognition
Zhang, HY; Han, ZB; Xu, B
Source PublicationCHINESE JOURNAL OF ELECTRONICS
2004-10-01
Volume13Issue:4Pages:665-670
SubtypeArticle
AbstractThe degradation of speech recognition performance in real-life environments and through transmission channels is a main embarrassment for many speech-based applications around the world, especially when non-stationary noise and changing channel exist. Previous works have shown that the main reason for this performance degradation is the variational mismatch caused by different telephone channels between the testing and training sets. In this paper, we propose a statistic model based implementation to dynamically compensate this mismatch. Firstly, we focus on a Maximum-likelihood (ML) estimation algorithm for telephone channels. In experiments on Mandarin Large vocabulary continuous speech recognition (LVCSR) over telephone lines, the Character error rate (CER) decreases more than 20%. The average delay is about 300similar to400ms. Secondly, we will extend it by introducing a phone-conditioned prior statistic model for the channels and applying Maximum a posteriori (MAP) estimation technique. Compared to the ML based method, the MAP based algorithm follows with the variations within channels more effectively. Average delay of the algorithm is decreased to 200ms. An additional 7similar to8% CER relative reduction is observed in LVCSR.
KeywordAutomatic Speech Recognition (Asr) Telephone Channel Compensation Statistic Model Maximum Likelihood Estimation Maximum a Posteriori Estimation
WOS HeadingsScience & Technology ; Technology
Indexed BySCI
Language英语
WOS Research AreaEngineering
WOS SubjectEngineering, Electrical & Electronic
WOS IDWOS:000224787200024
Citation statistics
Document Type期刊论文
Identifierhttp://ir.ia.ac.cn/handle/173211/8927
Collection09年以前成果
Affiliation1.Chinese Acad Sci, Inst Automat, Hightech Innovat Ctr, Beijing 100080, Peoples R China
2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100080, Peoples R China
Recommended Citation
GB/T 7714
Zhang, HY,Han, ZB,Xu, B. Statistic model based dynamic channel compensation for telephony speech recognition[J]. CHINESE JOURNAL OF ELECTRONICS,2004,13(4):665-670.
APA Zhang, HY,Han, ZB,&Xu, B.(2004).Statistic model based dynamic channel compensation for telephony speech recognition.CHINESE JOURNAL OF ELECTRONICS,13(4),665-670.
MLA Zhang, HY,et al."Statistic model based dynamic channel compensation for telephony speech recognition".CHINESE JOURNAL OF ELECTRONICS 13.4(2004):665-670.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Zhang, HY]'s Articles
[Han, ZB]'s Articles
[Xu, B]'s Articles
Baidu academic
Similar articles in Baidu academic
[Zhang, HY]'s Articles
[Han, ZB]'s Articles
[Xu, B]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Zhang, HY]'s Articles
[Han, ZB]'s Articles
[Xu, B]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

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