GA-Based Optimization of Temporal Filter for Robust Speech Recognition
碩士 === 國立暨南國際大學 === 電機工程學系 === 95 === This thesis proposed a new robust speech recognition technique in noisy environment. The feature extraction bases on MFCC (Mel-frequency cepstral coefficients), and template matching employs Hidden Markov Models (HMM). Since the performance of speech recognition...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2006
|
Online Access: | http://ndltd.ncl.edu.tw/handle/14018764062129674426 |
id |
ndltd-TW-094NCNU0442038 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-094NCNU04420382015-10-13T10:38:05Z http://ndltd.ncl.edu.tw/handle/14018764062129674426 GA-Based Optimization of Temporal Filter for Robust Speech Recognition 基於基因遺傳演算法最佳時序濾波器之應用強健性語音辨識 Kuei-Ting Kuo éæ¡å»· 碩士 國立暨南國際大學 電機工程學系 95 This thesis proposed a new robust speech recognition technique in noisy environment. The feature extraction bases on MFCC (Mel-frequency cepstral coefficients), and template matching employs Hidden Markov Models (HMM). Since the performance of speech recognition can be improved by using temporal filters, we focus on the optimization of these filters. Hence, we adopt genetic algorithms (GA) to dynamically select the proper temporal filters in order to obtain the robust MFCC. For Mel-scale banks, there are totally 20 triangular banks. Hence, there are 20 corresponding temporal filters which are encoded into the chromosome. We use 10 chromosomes in the genetic population. Finally, we do the experiment, it adopt Chinese digit (0-9) words form 20 speakers. Everyone speaks 10 times. One half people speak as reference data, other as test data. The recognition rate can attain 44.5% in 0db SNR. Gin-Der Wu 吳俊德 2006 學位論文 ; thesis 38 en_US |
collection |
NDLTD |
language |
en_US |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立暨南國際大學 === 電機工程學系 === 95 === This thesis proposed a new robust speech recognition technique in noisy environment. The feature extraction bases on MFCC (Mel-frequency cepstral coefficients), and template matching employs Hidden Markov Models (HMM). Since the performance of speech recognition can be improved by using temporal filters, we focus on the optimization of these filters. Hence, we adopt genetic algorithms (GA) to dynamically select the proper temporal filters in order to obtain the robust MFCC. For Mel-scale banks, there are totally 20 triangular banks. Hence, there are 20 corresponding temporal filters which are encoded into the chromosome. We use 10 chromosomes in the genetic population. Finally, we do the experiment, it adopt Chinese digit (0-9) words form 20 speakers. Everyone speaks 10 times. One half people speak as reference data, other as test data. The recognition rate can attain 44.5% in 0db SNR.
|
author2 |
Gin-Der Wu |
author_facet |
Gin-Der Wu Kuei-Ting Kuo éæ¡å»· |
author |
Kuei-Ting Kuo éæ¡å»· |
spellingShingle |
Kuei-Ting Kuo éæ¡å»· GA-Based Optimization of Temporal Filter for Robust Speech Recognition |
author_sort |
Kuei-Ting Kuo |
title |
GA-Based Optimization of Temporal Filter for Robust Speech Recognition |
title_short |
GA-Based Optimization of Temporal Filter for Robust Speech Recognition |
title_full |
GA-Based Optimization of Temporal Filter for Robust Speech Recognition |
title_fullStr |
GA-Based Optimization of Temporal Filter for Robust Speech Recognition |
title_full_unstemmed |
GA-Based Optimization of Temporal Filter for Robust Speech Recognition |
title_sort |
ga-based optimization of temporal filter for robust speech recognition |
publishDate |
2006 |
url |
http://ndltd.ncl.edu.tw/handle/14018764062129674426 |
work_keys_str_mv |
AT kueitingkuo gabasedoptimizationoftemporalfilterforrobustspeechrecognition AT eæa gabasedoptimizationoftemporalfilterforrobustspeechrecognition AT kueitingkuo jīyújīyīnyíchuányǎnsuànfǎzuìjiāshíxùlǜbōqìzhīyīngyòngqiángjiànxìngyǔyīnbiànshí AT eæa jīyújīyīnyíchuányǎnsuànfǎzuìjiāshíxùlǜbōqìzhīyīngyòngqiángjiànxìngyǔyīnbiànshí |
_version_ |
1716831511856021504 |