基于HMM-GMM的Amazigh语音识别系统

IF 0.6 Q3 Engineering

International Journal of Signal and Imaging Systems Engineering Pub Date : 2020-01-01 DOI:10.1504/IJSISE.2020.10036146

Safâa El Ouahabi, M. Atounti, Mohamed Bellouki

{"title":"基于HMM-GMM的Amazigh语音识别系统","authors":"Safâa El Ouahabi, M. Atounti, Mohamed Bellouki","doi":"10.1504/IJSISE.2020.10036146","DOIUrl":null,"url":null,"abstract":"This study presents conception and realisation of an automatic independent speech recognition system using hidden Markov model (HMM). The system recognises 33 letters in Amazigh language. System is found well performed and can identify the Amazigh spoken letters at 88, 44% recognition rate, which is well acceptable rate of accuracy for speech recognition. The tests were taken based on the HMM and Gaussian mixture distributions. Hidden Markov toolkit (HTK) has been used in implementation and test phases. The word error rate (WER) came initially to 29.41 and reduced to about 11.52% thanks to extensive testing and change of the recognition's parameters.","PeriodicalId":56359,"journal":{"name":"International Journal of Signal and Imaging Systems Engineering","volume":"1 1","pages":""},"PeriodicalIF":0.6000,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"HMM-GMM based Amazigh speech recognition system\",\"authors\":\"Safâa El Ouahabi, M. Atounti, Mohamed Bellouki\",\"doi\":\"10.1504/IJSISE.2020.10036146\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This study presents conception and realisation of an automatic independent speech recognition system using hidden Markov model (HMM). The system recognises 33 letters in Amazigh language. System is found well performed and can identify the Amazigh spoken letters at 88, 44% recognition rate, which is well acceptable rate of accuracy for speech recognition. The tests were taken based on the HMM and Gaussian mixture distributions. Hidden Markov toolkit (HTK) has been used in implementation and test phases. The word error rate (WER) came initially to 29.41 and reduced to about 11.52% thanks to extensive testing and change of the recognition's parameters.\",\"PeriodicalId\":56359,\"journal\":{\"name\":\"International Journal of Signal and Imaging Systems Engineering\",\"volume\":\"1 1\",\"pages\":\"\"},\"PeriodicalIF\":0.6000,\"publicationDate\":\"2020-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"International Journal of Signal and Imaging Systems Engineering\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1504/IJSISE.2020.10036146\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"Engineering\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Signal and Imaging Systems Engineering","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1504/IJSISE.2020.10036146","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"Engineering","Score":null,"Total":0}

引用次数: 2

摘要

本文提出了一种基于隐马尔可夫模型的自动独立语音识别系统的概念和实现。该系统可以识别阿马齐格语中的33个字母。结果表明，该系统具有良好的性能，能够以88.44%的识别率识别Amazigh语音字母，这是语音识别可以接受的准确率。基于HMM和高斯混合分布进行了测试。隐马尔可夫工具包(HTK)已在实现和测试阶段使用。单词错误率(WER)最初为29.41，由于大量的测试和识别参数的改变，降低到11.52%左右。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

HMM-GMM based Amazigh speech recognition system

This study presents conception and realisation of an automatic independent speech recognition system using hidden Markov model (HMM). The system recognises 33 letters in Amazigh language. System is found well performed and can identify the Amazigh spoken letters at 88, 44% recognition rate, which is well acceptable rate of accuracy for speech recognition. The tests were taken based on the HMM and Gaussian mixture distributions. Hidden Markov toolkit (HTK) has been used in implementation and test phases. The word error rate (WER) came initially to 29.41 and reduced to about 11.52% thanks to extensive testing and change of the recognition's parameters.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

International Journal of Signal and Imaging Systems Engineering ENGINEERING, ELECTRICAL & ELECTRONIC-

CiteScore

2.10

自引率

0.00%

发文量