Return to search

A design of speech recognition system for one hundred thousand Chinese names

The objective of this thesis is to design and implement a speech recognition system for one hundred thousand Chinese names. Mel frequency cepstrum coefficient, hidden Markov model and lexicon search strategy are utilized to choose the name candidates. Furthermore, a mandarin intonation technique is also incorporated into this system to increase the final speech recognition accuracy.
The experimental results indicate that for the speaker dependent case, an 85% correct rate can be achieved by use of the proposed intonation classification scheme and the balanced monosyllable training database. The above correct rate has an increase of 8% over the previous method without using these two techniques. Under Redhat Linux 9.0 environment, a mandarin name can be recognized within 2 seconds by the use of a computer with Intel Celeron 2.4 GHz CPU.

Identiferoai:union.ndltd.org:NSYSU/oai:NSYSU:etd-0906107-041006
Date06 September 2007
CreatorsTu, Chiu-chuan
ContributorsTsung Lee, Chih-Chien Chen, Chii Maw Uang
PublisherNSYSU
Source SetsNSYSU Electronic Thesis and Dissertation Archive
LanguageCholon
Detected LanguageEnglish
Typetext
Formatapplication/pdf
Sourcehttp://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0906107-041006
Rightsnot_available, Copyright information available at source archive

Page generated in 0.0021 seconds