Cover Image

Speaker identification using distributed vector quantization and Gaussian mixture models

Speaker identification is the computing task of recognizing people's identity based on their voices. There are two main difficulties in this field. First is how to maintain the accuracy rate under large amount of training data. Second is how to reduce the processing time. Previous studies repor...

Full description

Saved in:
Bibliographic Details
Main Author: Loh, Mun Yee
Format: Thesis
Language:English
Published: 2010
Subjects:
QA75 Electronic computers. Computer science
Online Access:http://eprints.utm.my/id/eprint/11585/6/LohMunYeeMFSKSM2010.pdf
http://eprints.utm.my/id/eprint/11585/
Tags: Add Tag
No Tags, Be the first to tag this record!
id my.utm.11585
record_format eprints
spelling my.utm.115852017-09-28T00:21:43Z http://eprints.utm.my/id/eprint/11585/ Speaker identification using distributed vector quantization and Gaussian mixture models Loh, Mun Yee QA75 Electronic computers. Computer science Speaker identification is the computing task of recognizing people's identity based on their voices. There are two main difficulties in this field. First is how to maintain the accuracy rate under large amount of training data. Second is how to reduce the processing time. Previous studies reported that Gaussian Mixture Model (GMM) for speaker identification appears to have many advantages. However, due to long processing time, this process does not always produce satisfying result in practice. Meanwhile, current mechanisms for hybrid production of speaker identification are directed more towards accuracy problems, not processing time optimization. This research focuses on constructing distributed data training on Vector Quantization (VQ) modeling to achieve an initial result. The decision tree approach is applied to obtain distributed training for VQ model. GMM classification process is then employed on the initial result to achieve a final result. The efficiency of the model is evaluated by computational time and accuracy rate compared to GMM baseline models. Experimental result shows that the hybrid distributed VQ/GMM model yields better accuracy. Besides, it gives 80% reduction in processing time and is 5 times faster compared to GMM baseline models. In conclusion, this research successfully improves the computational time and accuracy of the text-independent speaker identification system. 2010-03 Thesis NonPeerReviewed application/pdf en http://eprints.utm.my/id/eprint/11585/6/LohMunYeeMFSKSM2010.pdf Loh, Mun Yee (2010) Speaker identification using distributed vector quantization and Gaussian mixture models. Masters thesis, Universiti Teknologi Malaysia, Faculty of Computer Science and Information Systems.
institution Universiti Teknologi Malaysia
building UTM Library
collection Institutional Repository
continent Asia
country Malaysia
content_provider Universiti Teknologi Malaysia
content_source UTM Institutional Repository
url_provider http://eprints.utm.my/
language English
topic QA75 Electronic computers. Computer science
spellingShingle QA75 Electronic computers. Computer science
Loh, Mun Yee
Speaker identification using distributed vector quantization and Gaussian mixture models
description Speaker identification is the computing task of recognizing people's identity based on their voices. There are two main difficulties in this field. First is how to maintain the accuracy rate under large amount of training data. Second is how to reduce the processing time. Previous studies reported that Gaussian Mixture Model (GMM) for speaker identification appears to have many advantages. However, due to long processing time, this process does not always produce satisfying result in practice. Meanwhile, current mechanisms for hybrid production of speaker identification are directed more towards accuracy problems, not processing time optimization. This research focuses on constructing distributed data training on Vector Quantization (VQ) modeling to achieve an initial result. The decision tree approach is applied to obtain distributed training for VQ model. GMM classification process is then employed on the initial result to achieve a final result. The efficiency of the model is evaluated by computational time and accuracy rate compared to GMM baseline models. Experimental result shows that the hybrid distributed VQ/GMM model yields better accuracy. Besides, it gives 80% reduction in processing time and is 5 times faster compared to GMM baseline models. In conclusion, this research successfully improves the computational time and accuracy of the text-independent speaker identification system.
format Thesis
author Loh, Mun Yee
author_facet Loh, Mun Yee
author_sort Loh, Mun Yee
title Speaker identification using distributed vector quantization and Gaussian mixture models
title_short Speaker identification using distributed vector quantization and Gaussian mixture models
title_full Speaker identification using distributed vector quantization and Gaussian mixture models
title_fullStr Speaker identification using distributed vector quantization and Gaussian mixture models
title_full_unstemmed Speaker identification using distributed vector quantization and Gaussian mixture models
title_sort speaker identification using distributed vector quantization and gaussian mixture models
publishDate 2010
url http://eprints.utm.my/id/eprint/11585/6/LohMunYeeMFSKSM2010.pdf
http://eprints.utm.my/id/eprint/11585/
_version_ 1643645722561609728
score 13.211869

Similar Items