Implementation and analysis of GMM-based speaker identification on FPGA

The use of highly accurate identification systems is required in today’s society. Existing systems such as pin numbers and passwords can be forgotten or forged easily and they are no longer considered to offer a high level of security. The use of biological features (biometrics) is becoming widely a...

Full description

Saved in:

Bibliographic Details
Main Author:	Phaklen, Ehkan
Format:	Thesis
Language:	English
Subjects:	Field Programmable Gate Array (FPGA) Gaussian Mixture Model (GMM) Personal identification systems Biometric-based speaker identification Speech signals
Online Access:	http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/32469/1/Page%201-24.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/32469/2/Full%20text.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my-unimap-32469
record_format	uketd_dc
spelling	my-unimap-324692014-03-09T12:53:43Z Implementation and analysis of GMM-based speaker identification on FPGA Phaklen, Ehkan The use of highly accurate identification systems is required in today’s society. Existing systems such as pin numbers and passwords can be forgotten or forged easily and they are no longer considered to offer a high level of security. The use of biological features (biometrics) is becoming widely accepted as the next level for security systems. One of the biometric is the human voice and it leads to the task of speaker identification. Speaker identification is the process of determining whether a speaker exists in a group of known speakers and identifying the speaker within the group. Speaker specific characteristics exist in speech signals due to different speakers having different resonances of the vocal tract. These differences can be exploited by extracting Mel-frequency Cepstral Coefficients (MFCCs) from the speech signal. A statistical modelling process known as Gaussian Mixture Model (GMM) is used to model the distribution of each speaker’s MFCCs in a multi-dimensional acoustic space. GMM involves with two phases called training and classification. The training phase is complex and is better suited for implementation in software. The classification phase is well suited for implementation in hardware and this allows for real time processing of multiple voice streams on large population sizes. Several innovative techniques are demonstrated which enable hardware system to obtain two orders of magnitude speed up over software while maintaining comparable levels of accuracy. A speedup factor of eighty six is achieved on hardware-based FPGA compared to a software implementation on a standard PC for this approach. Universiti Malaysia Perlis (UniMAP) 2012 Thesis en http://dspace.unimap.edu.my:80/dspace/handle/123456789/32469 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/32469/1/Page%201-24.pdf 74856b62a34a9d54940f531e3c1cf4d2 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/32469/2/Full%20text.pdf 4865af4bc6bbd494ccfc9f124d420625 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/32469/3/license.txt 8a4605be74aa9ea9d79846c1fba20a33 Field Programmable Gate Array (FPGA) Gaussian Mixture Model (GMM) Personal identification systems Biometric-based speaker identification Speech signals School of Computer and Communication Engineering
institution	Universiti Malaysia Perlis
collection	UniMAP Institutional Repository
language	English
topic	Field Programmable Gate Array (FPGA) Gaussian Mixture Model (GMM) Personal identification systems Biometric-based speaker identification Speech signals
spellingShingle	Field Programmable Gate Array (FPGA) Gaussian Mixture Model (GMM) Personal identification systems Biometric-based speaker identification Speech signals Phaklen, Ehkan Implementation and analysis of GMM-based speaker identification on FPGA
description	The use of highly accurate identification systems is required in today’s society. Existing systems such as pin numbers and passwords can be forgotten or forged easily and they are no longer considered to offer a high level of security. The use of biological features (biometrics) is becoming widely accepted as the next level for security systems. One of the biometric is the human voice and it leads to the task of speaker identification. Speaker identification is the process of determining whether a speaker exists in a group of known speakers and identifying the speaker within the group. Speaker specific characteristics exist in speech signals due to different speakers having different resonances of the vocal tract. These differences can be exploited by extracting Mel-frequency Cepstral Coefficients (MFCCs) from the speech signal. A statistical modelling process known as Gaussian Mixture Model (GMM) is used to model the distribution of each speaker’s MFCCs in a multi-dimensional acoustic space. GMM involves with two phases called training and classification. The training phase is complex and is better suited for implementation in software. The classification phase is well suited for implementation in hardware and this allows for real time processing of multiple voice streams on large population sizes. Several innovative techniques are demonstrated which enable hardware system to obtain two orders of magnitude speed up over software while maintaining comparable levels of accuracy. A speedup factor of eighty six is achieved on hardware-based FPGA compared to a software implementation on a standard PC for this approach.
format	Thesis
author	Phaklen, Ehkan
author_facet	Phaklen, Ehkan
author_sort	Phaklen, Ehkan
title	Implementation and analysis of GMM-based speaker identification on FPGA
title_short	Implementation and analysis of GMM-based speaker identification on FPGA
title_full	Implementation and analysis of GMM-based speaker identification on FPGA
title_fullStr	Implementation and analysis of GMM-based speaker identification on FPGA
title_full_unstemmed	Implementation and analysis of GMM-based speaker identification on FPGA
title_sort	implementation and analysis of gmm-based speaker identification on fpga
granting_institution	Universiti Malaysia Perlis (UniMAP)
granting_department	School of Computer and Communication Engineering
url	http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/32469/1/Page%201-24.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/32469/2/Full%20text.pdf
_version_	1747836801608843264

Implementation and analysis of GMM-based speaker identification on FPGA

Similar Items