A Design of German Speech Recognition System
碩士 === 國立中山大學 === 電機工程學系研究所 === 98 === This thesis investigates the design and implementation strategies for a German speech recognition system. It utilizes the speech features of the 434 common German mono-syllables as the major training and recognition methodology. A training database is establish...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2010
|
Online Access: | http://ndltd.ncl.edu.tw/handle/86665595218199048931 |
id |
ndltd-TW-098NSYS5442093 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-098NSYS54420932015-10-13T18:39:46Z http://ndltd.ncl.edu.tw/handle/86665595218199048931 A Design of German Speech Recognition System 德文語音辨識系統之設計研究 Shih-Sin Lai 賴世欣 碩士 國立中山大學 電機工程學系研究所 98 This thesis investigates the design and implementation strategies for a German speech recognition system. It utilizes the speech features of the 434 common German mono-syllables as the major training and recognition methodology. A training database is established by reading each mono-syllable 12 times in 6 rounds. Every mono-syllable is consecutively read twice with different tones. The first pronounced pattern has high pitch of tone 1, while the second one has falling pitch of tone 4. Mel-frequency cepstral coefficients, linear predictive cepstral coefficients, and hidden Markov model are used as the two feature models and the recognition model respectively. Under the AMD Athlon X2-240 with 2.8 GHz clock rate personal computer and Ubuntu 9.04 operating system environment, a correct phrase recognition rate of 84% can be reached for a 3900 German phrase database. The average computation time for each phrase is within 1 second. Chih-Chien Chen 陳志堅 2010 學位論文 ; thesis 50 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立中山大學 === 電機工程學系研究所 === 98 === This thesis investigates the design and implementation strategies for a German speech recognition system. It utilizes the speech features of the 434 common German mono-syllables as the major training and recognition methodology. A training database is established by reading each mono-syllable 12 times in 6 rounds. Every mono-syllable is consecutively read twice with different tones. The first pronounced pattern has high pitch of tone 1, while the second one has falling pitch of tone 4. Mel-frequency cepstral coefficients, linear predictive cepstral coefficients, and hidden Markov model are used as the two feature models and the recognition model respectively. Under the AMD Athlon X2-240 with 2.8 GHz clock rate personal computer and Ubuntu 9.04 operating system environment, a correct phrase recognition rate of 84% can be reached for a 3900 German phrase database. The average computation time for each phrase is within 1 second.
|
author2 |
Chih-Chien Chen |
author_facet |
Chih-Chien Chen Shih-Sin Lai 賴世欣 |
author |
Shih-Sin Lai 賴世欣 |
spellingShingle |
Shih-Sin Lai 賴世欣 A Design of German Speech Recognition System |
author_sort |
Shih-Sin Lai |
title |
A Design of German Speech Recognition System |
title_short |
A Design of German Speech Recognition System |
title_full |
A Design of German Speech Recognition System |
title_fullStr |
A Design of German Speech Recognition System |
title_full_unstemmed |
A Design of German Speech Recognition System |
title_sort |
design of german speech recognition system |
publishDate |
2010 |
url |
http://ndltd.ncl.edu.tw/handle/86665595218199048931 |
work_keys_str_mv |
AT shihsinlai adesignofgermanspeechrecognitionsystem AT làishìxīn adesignofgermanspeechrecognitionsystem AT shihsinlai déwényǔyīnbiànshíxìtǒngzhīshèjìyánjiū AT làishìxīn déwényǔyīnbiànshíxìtǒngzhīshèjìyánjiū AT shihsinlai designofgermanspeechrecognitionsystem AT làishìxīn designofgermanspeechrecognitionsystem |
_version_ |
1718036497058234368 |