Speaker Identification Using Discriminative Features and Sparse Representation

Yu Hao Chin, Jia Ching Wang, Chien Lin Huang, Kuang Yao Wang, Chung-Hsien Wu

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Speaker identification is an important topic with relevance to various disciplines. This paper proposes a novel speaker identification system, which consists of two major components-feature extraction and sparse representation classifier (SRC). Although SRC has been utilized for many classification purposes, few studies have provided insight into the link between the commonly used speaker identification feature, i-vector, and SRC. To combine i-vector and SRC sufficiently, we use probabilistic principal component analysis and Bartlett test to extract high-quality i-vector to construct a discriminative dictionary in SRC, supporting effective speaker identification. Besides improving dictionary from the i-vector aspect, we also utilize dictionary learning to further enhance the content of the dictionary. Two learning methods are proposed-robust principal component analysis dictionary and SVD-dictionary. Furthermore, we propose constructing a noise dictionary and combine it with the original dictionary to absorb and suppress noise when implementing the sparse coding. Various coding methods are utilized and analyzed. A comparison to the methods for speaker identification reveals that the proposed method outperforms the baselines and confirms its feasibility.

Original languageEnglish
Article number7872470
Pages (from-to)1979-1987
Number of pages9
JournalIEEE Transactions on Information Forensics and Security
Volume12
Issue number8
DOIs
Publication statusPublished - 2017 Aug 1

Fingerprint

Glossaries
Classifiers
Principal component analysis
Singular value decomposition
Feature extraction
Identification (control systems)

All Science Journal Classification (ASJC) codes

  • Safety, Risk, Reliability and Quality
  • Computer Networks and Communications

Cite this

Chin, Yu Hao ; Wang, Jia Ching ; Huang, Chien Lin ; Wang, Kuang Yao ; Wu, Chung-Hsien. / Speaker Identification Using Discriminative Features and Sparse Representation. In: IEEE Transactions on Information Forensics and Security. 2017 ; Vol. 12, No. 8. pp. 1979-1987.
@article{9dc7cf7141594d838e6250886975bf23,
title = "Speaker Identification Using Discriminative Features and Sparse Representation",
abstract = "Speaker identification is an important topic with relevance to various disciplines. This paper proposes a novel speaker identification system, which consists of two major components-feature extraction and sparse representation classifier (SRC). Although SRC has been utilized for many classification purposes, few studies have provided insight into the link between the commonly used speaker identification feature, i-vector, and SRC. To combine i-vector and SRC sufficiently, we use probabilistic principal component analysis and Bartlett test to extract high-quality i-vector to construct a discriminative dictionary in SRC, supporting effective speaker identification. Besides improving dictionary from the i-vector aspect, we also utilize dictionary learning to further enhance the content of the dictionary. Two learning methods are proposed-robust principal component analysis dictionary and SVD-dictionary. Furthermore, we propose constructing a noise dictionary and combine it with the original dictionary to absorb and suppress noise when implementing the sparse coding. Various coding methods are utilized and analyzed. A comparison to the methods for speaker identification reveals that the proposed method outperforms the baselines and confirms its feasibility.",
author = "Chin, {Yu Hao} and Wang, {Jia Ching} and Huang, {Chien Lin} and Wang, {Kuang Yao} and Chung-Hsien Wu",
year = "2017",
month = "8",
day = "1",
doi = "10.1109/TIFS.2017.2678458",
language = "English",
volume = "12",
pages = "1979--1987",
journal = "IEEE Transactions on Information Forensics and Security",
issn = "1556-6013",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "8",

}

Speaker Identification Using Discriminative Features and Sparse Representation. / Chin, Yu Hao; Wang, Jia Ching; Huang, Chien Lin; Wang, Kuang Yao; Wu, Chung-Hsien.

In: IEEE Transactions on Information Forensics and Security, Vol. 12, No. 8, 7872470, 01.08.2017, p. 1979-1987.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Speaker Identification Using Discriminative Features and Sparse Representation

AU - Chin, Yu Hao

AU - Wang, Jia Ching

AU - Huang, Chien Lin

AU - Wang, Kuang Yao

AU - Wu, Chung-Hsien

PY - 2017/8/1

Y1 - 2017/8/1

N2 - Speaker identification is an important topic with relevance to various disciplines. This paper proposes a novel speaker identification system, which consists of two major components-feature extraction and sparse representation classifier (SRC). Although SRC has been utilized for many classification purposes, few studies have provided insight into the link between the commonly used speaker identification feature, i-vector, and SRC. To combine i-vector and SRC sufficiently, we use probabilistic principal component analysis and Bartlett test to extract high-quality i-vector to construct a discriminative dictionary in SRC, supporting effective speaker identification. Besides improving dictionary from the i-vector aspect, we also utilize dictionary learning to further enhance the content of the dictionary. Two learning methods are proposed-robust principal component analysis dictionary and SVD-dictionary. Furthermore, we propose constructing a noise dictionary and combine it with the original dictionary to absorb and suppress noise when implementing the sparse coding. Various coding methods are utilized and analyzed. A comparison to the methods for speaker identification reveals that the proposed method outperforms the baselines and confirms its feasibility.

AB - Speaker identification is an important topic with relevance to various disciplines. This paper proposes a novel speaker identification system, which consists of two major components-feature extraction and sparse representation classifier (SRC). Although SRC has been utilized for many classification purposes, few studies have provided insight into the link between the commonly used speaker identification feature, i-vector, and SRC. To combine i-vector and SRC sufficiently, we use probabilistic principal component analysis and Bartlett test to extract high-quality i-vector to construct a discriminative dictionary in SRC, supporting effective speaker identification. Besides improving dictionary from the i-vector aspect, we also utilize dictionary learning to further enhance the content of the dictionary. Two learning methods are proposed-robust principal component analysis dictionary and SVD-dictionary. Furthermore, we propose constructing a noise dictionary and combine it with the original dictionary to absorb and suppress noise when implementing the sparse coding. Various coding methods are utilized and analyzed. A comparison to the methods for speaker identification reveals that the proposed method outperforms the baselines and confirms its feasibility.

UR - http://www.scopus.com/inward/record.url?scp=85020267724&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020267724&partnerID=8YFLogxK

U2 - 10.1109/TIFS.2017.2678458

DO - 10.1109/TIFS.2017.2678458

M3 - Article

VL - 12

SP - 1979

EP - 1987

JO - IEEE Transactions on Information Forensics and Security

JF - IEEE Transactions on Information Forensics and Security

SN - 1556-6013

IS - 8

M1 - 7872470

ER -