Wavelet-domain hidden Markov models for signal detection and classification

M. S. Crouse; R. D. Nowak; Kerim Mhirsi; Richb G. Baraniuk

doi:10.1117/12.279500

Wavelet-domain hidden Markov models for signal detection and classification

M. S. Crouse, R. D. Nowak, Kerim Mhirsi, Richb G. Baraniuk

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Scopus citations

Abstract

This paper addresses the problem of detection and classification of complicated signals in noise. Classical detection methods such as energy detectors and linear discriminant analysis do not perform well in many situations of practical interest. We introduce a new approach based on hidden Markov modeling in the wavelet domain. Using training data, we fit a hidden Markov model (HMM) to the wavelet transform to concisely represent its probabilistic time-frequency structure. The HMM provides a natural framework for performing likelihood ratio tests used in signal detection and classification. We compare our approach with classical methods for classification of nonlinear processes, change-point detection, and detection with unknown delay.

Original language	English (US)
Title of host publication	Proceedings of SPIE - The International Society for Optical Engineering
Editors	F.T. Luk
Pages	36-47
Number of pages	12
Volume	3162
DOIs	https://doi.org/10.1117/12.279500
State	Published - 1997
Event	Advanced Signal Processing: Algorithms, Architectures and Implementations VII - San Diego, CA, United States Duration: Jul 28 1997 → Jul 30 1997

Other

Other	Advanced Signal Processing: Algorithms, Architectures and Implementations VII
Country/Territory	United States
City	San Diego, CA
Period	7/28/97 → 7/30/97

Keywords

Classification
Detection
Hidden Markov models
Wavelets

ASJC Scopus subject areas

Electrical and Electronic Engineering
Condensed Matter Physics

Access to Document

10.1117/12.279500

Cite this

Crouse, MS, Nowak, RD, Mhirsi, K & Baraniuk, RG 1997, Wavelet-domain hidden Markov models for signal detection and classification. in FT Luk (ed.), Proceedings of SPIE - The International Society for Optical Engineering. vol. 3162, pp. 36-47, Advanced Signal Processing: Algorithms, Architectures and Implementations VII, San Diego, CA, United States, 7/28/97. https://doi.org/10.1117/12.279500

@inproceedings{50f6452b613a42e2b9c085b767f809f8,

title = "Wavelet-domain hidden Markov models for signal detection and classification",

abstract = "This paper addresses the problem of detection and classification of complicated signals in noise. Classical detection methods such as energy detectors and linear discriminant analysis do not perform well in many situations of practical interest. We introduce a new approach based on hidden Markov modeling in the wavelet domain. Using training data, we fit a hidden Markov model (HMM) to the wavelet transform to concisely represent its probabilistic time-frequency structure. The HMM provides a natural framework for performing likelihood ratio tests used in signal detection and classification. We compare our approach with classical methods for classification of nonlinear processes, change-point detection, and detection with unknown delay.",

keywords = "Classification, Detection, Hidden Markov models, Wavelets",

author = "Crouse, {M. S.} and Nowak, {R. D.} and Kerim Mhirsi and Baraniuk, {Richb G.}",

year = "1997",

doi = "10.1117/12.279500",

language = "English (US)",

volume = "3162",

pages = "36--47",

editor = "F.T. Luk",

booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

note = "Advanced Signal Processing: Algorithms, Architectures and Implementations VII ; Conference date: 28-07-1997 Through 30-07-1997",

}

TY - GEN

T1 - Wavelet-domain hidden Markov models for signal detection and classification

AU - Crouse, M. S.

AU - Nowak, R. D.

AU - Mhirsi, Kerim

AU - Baraniuk, Richb G.

PY - 1997

Y1 - 1997

N2 - This paper addresses the problem of detection and classification of complicated signals in noise. Classical detection methods such as energy detectors and linear discriminant analysis do not perform well in many situations of practical interest. We introduce a new approach based on hidden Markov modeling in the wavelet domain. Using training data, we fit a hidden Markov model (HMM) to the wavelet transform to concisely represent its probabilistic time-frequency structure. The HMM provides a natural framework for performing likelihood ratio tests used in signal detection and classification. We compare our approach with classical methods for classification of nonlinear processes, change-point detection, and detection with unknown delay.

AB - This paper addresses the problem of detection and classification of complicated signals in noise. Classical detection methods such as energy detectors and linear discriminant analysis do not perform well in many situations of practical interest. We introduce a new approach based on hidden Markov modeling in the wavelet domain. Using training data, we fit a hidden Markov model (HMM) to the wavelet transform to concisely represent its probabilistic time-frequency structure. The HMM provides a natural framework for performing likelihood ratio tests used in signal detection and classification. We compare our approach with classical methods for classification of nonlinear processes, change-point detection, and detection with unknown delay.

KW - Classification

KW - Detection

KW - Hidden Markov models

KW - Wavelets

UR - http://www.scopus.com/inward/record.url?scp=0031290048&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031290048&partnerID=8YFLogxK

U2 - 10.1117/12.279500

DO - 10.1117/12.279500

M3 - Conference contribution

AN - SCOPUS:0031290048

VL - 3162

SP - 36

EP - 47

BT - Proceedings of SPIE - The International Society for Optical Engineering

A2 - Luk, F.T.

T2 - Advanced Signal Processing: Algorithms, Architectures and Implementations VII

Y2 - 28 July 1997 through 30 July 1997

ER -

Wavelet-domain hidden Markov models for signal detection and classification

Abstract

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this