Face recognition by sparse representation

Arvind Ganesh, Andrew Wagner, Zihan Zhou, Allen Y. Yang, Yi Ma, John Wright

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

In this chapter, we present a comprehensive framework for tackling the classical problem of face recognition, based on theory and algorithms from sparse representation. Despite intense interest in the past several decades, traditional pattern recognition theory still stops short of providing a satisfactory solution capable of recognizing human faces in the presence of real-world nuisances such as occlusion and variabilities in pose and illumination. Our new approach, called sparse representation-based classification (SRC), is motivated by a very natural notion of sparsity, namely, one should always try to explain a query image using a small number of training images from a single subject category. This sparse representation is sought via ℓ1 minimization. We show how this core idea can be generalized and extended to account for various physical variabilities encountered in face recognition. The end result of our investigation is a full-fledged practical system aimed at security and access control applications. The system is capable of accurately recognizing subjects out of a database of several hundred subjects with state-of-the-art accuracy. Introduction. Automatic face recognition is a classical problem in the computer vision community. The communityʼs sustained interest in this problem is mainly due to two reasons. First, in face recognition, we encounter many of the common variabilities that plague vision systems in general: illumination, occlusion, pose, and misalignment. Inspired by the good performance of humans in recognizing familiar faces [38], we have reason to believe that effective automatic face recognition is possible, and that the quest to achieve this will tell us something about visual recognition in general.

Original languageEnglish (US)
Title of host publicationCompressed Sensing
Subtitle of host publicationTheory and Applications
PublisherCambridge University Press
Pages515-539
Number of pages25
ISBN (Electronic)9780511794308
ISBN (Print)9781107005587
DOIs
StatePublished - Jan 1 2009

Fingerprint

Face recognition
Lighting
Access control
Computer vision
Pattern recognition

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Ganesh, A., Wagner, A., Zhou, Z., Yang, A. Y., Ma, Y., & Wright, J. (2009). Face recognition by sparse representation. In Compressed Sensing: Theory and Applications (pp. 515-539). Cambridge University Press. https://doi.org/10.1017/CBO9780511794308.013
Ganesh, Arvind ; Wagner, Andrew ; Zhou, Zihan ; Yang, Allen Y. ; Ma, Yi ; Wright, John. / Face recognition by sparse representation. Compressed Sensing: Theory and Applications. Cambridge University Press, 2009. pp. 515-539
@inbook{58c371d5cc0c4e0d93fd9cf3208ca633,
title = "Face recognition by sparse representation",
abstract = "In this chapter, we present a comprehensive framework for tackling the classical problem of face recognition, based on theory and algorithms from sparse representation. Despite intense interest in the past several decades, traditional pattern recognition theory still stops short of providing a satisfactory solution capable of recognizing human faces in the presence of real-world nuisances such as occlusion and variabilities in pose and illumination. Our new approach, called sparse representation-based classification (SRC), is motivated by a very natural notion of sparsity, namely, one should always try to explain a query image using a small number of training images from a single subject category. This sparse representation is sought via ℓ1 minimization. We show how this core idea can be generalized and extended to account for various physical variabilities encountered in face recognition. The end result of our investigation is a full-fledged practical system aimed at security and access control applications. The system is capable of accurately recognizing subjects out of a database of several hundred subjects with state-of-the-art accuracy. Introduction. Automatic face recognition is a classical problem in the computer vision community. The communityʼs sustained interest in this problem is mainly due to two reasons. First, in face recognition, we encounter many of the common variabilities that plague vision systems in general: illumination, occlusion, pose, and misalignment. Inspired by the good performance of humans in recognizing familiar faces [38], we have reason to believe that effective automatic face recognition is possible, and that the quest to achieve this will tell us something about visual recognition in general.",
author = "Arvind Ganesh and Andrew Wagner and Zihan Zhou and Yang, {Allen Y.} and Yi Ma and John Wright",
year = "2009",
month = "1",
day = "1",
doi = "10.1017/CBO9780511794308.013",
language = "English (US)",
isbn = "9781107005587",
pages = "515--539",
booktitle = "Compressed Sensing",
publisher = "Cambridge University Press",
address = "United Kingdom",

}

Ganesh, A, Wagner, A, Zhou, Z, Yang, AY, Ma, Y & Wright, J 2009, Face recognition by sparse representation. in Compressed Sensing: Theory and Applications. Cambridge University Press, pp. 515-539. https://doi.org/10.1017/CBO9780511794308.013

Face recognition by sparse representation. / Ganesh, Arvind; Wagner, Andrew; Zhou, Zihan; Yang, Allen Y.; Ma, Yi; Wright, John.

Compressed Sensing: Theory and Applications. Cambridge University Press, 2009. p. 515-539.

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - Face recognition by sparse representation

AU - Ganesh, Arvind

AU - Wagner, Andrew

AU - Zhou, Zihan

AU - Yang, Allen Y.

AU - Ma, Yi

AU - Wright, John

PY - 2009/1/1

Y1 - 2009/1/1

N2 - In this chapter, we present a comprehensive framework for tackling the classical problem of face recognition, based on theory and algorithms from sparse representation. Despite intense interest in the past several decades, traditional pattern recognition theory still stops short of providing a satisfactory solution capable of recognizing human faces in the presence of real-world nuisances such as occlusion and variabilities in pose and illumination. Our new approach, called sparse representation-based classification (SRC), is motivated by a very natural notion of sparsity, namely, one should always try to explain a query image using a small number of training images from a single subject category. This sparse representation is sought via ℓ1 minimization. We show how this core idea can be generalized and extended to account for various physical variabilities encountered in face recognition. The end result of our investigation is a full-fledged practical system aimed at security and access control applications. The system is capable of accurately recognizing subjects out of a database of several hundred subjects with state-of-the-art accuracy. Introduction. Automatic face recognition is a classical problem in the computer vision community. The communityʼs sustained interest in this problem is mainly due to two reasons. First, in face recognition, we encounter many of the common variabilities that plague vision systems in general: illumination, occlusion, pose, and misalignment. Inspired by the good performance of humans in recognizing familiar faces [38], we have reason to believe that effective automatic face recognition is possible, and that the quest to achieve this will tell us something about visual recognition in general.

AB - In this chapter, we present a comprehensive framework for tackling the classical problem of face recognition, based on theory and algorithms from sparse representation. Despite intense interest in the past several decades, traditional pattern recognition theory still stops short of providing a satisfactory solution capable of recognizing human faces in the presence of real-world nuisances such as occlusion and variabilities in pose and illumination. Our new approach, called sparse representation-based classification (SRC), is motivated by a very natural notion of sparsity, namely, one should always try to explain a query image using a small number of training images from a single subject category. This sparse representation is sought via ℓ1 minimization. We show how this core idea can be generalized and extended to account for various physical variabilities encountered in face recognition. The end result of our investigation is a full-fledged practical system aimed at security and access control applications. The system is capable of accurately recognizing subjects out of a database of several hundred subjects with state-of-the-art accuracy. Introduction. Automatic face recognition is a classical problem in the computer vision community. The communityʼs sustained interest in this problem is mainly due to two reasons. First, in face recognition, we encounter many of the common variabilities that plague vision systems in general: illumination, occlusion, pose, and misalignment. Inspired by the good performance of humans in recognizing familiar faces [38], we have reason to believe that effective automatic face recognition is possible, and that the quest to achieve this will tell us something about visual recognition in general.

UR - http://www.scopus.com/inward/record.url?scp=84938545581&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84938545581&partnerID=8YFLogxK

U2 - 10.1017/CBO9780511794308.013

DO - 10.1017/CBO9780511794308.013

M3 - Chapter

AN - SCOPUS:84938545581

SN - 9781107005587

SP - 515

EP - 539

BT - Compressed Sensing

PB - Cambridge University Press

ER -

Ganesh A, Wagner A, Zhou Z, Yang AY, Ma Y, Wright J. Face recognition by sparse representation. In Compressed Sensing: Theory and Applications. Cambridge University Press. 2009. p. 515-539 https://doi.org/10.1017/CBO9780511794308.013