Face detection algorithm and feature performance on FRGC 2.0 imagery

J. Ross Beveridge; Andres Alvarez; Jilmil Saraf; Ward Fisher; Patrick J. Flynn; James Gentile

doi:10.1109/BTAS.2007.4401950

Face detection algorithm and feature performance on FRGC 2.0 imagery

J. Ross Beveridge, Andres Alvarez, Jilmil Saraf, Ward Fisher, Patrick J. Flynn, James Gentile

Earth Observation and Data Services

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

3 Scopus citations

Abstract

The performance of three well known face detection algorithms and four alternative types of features are characterized using face data from the Face Recognition Grand Challenge. The three algorithms are a Semi-Naive Bayesian Classifier, a neural network called a SNoW, and a Cascade Classifier using Haar wavelets. For the first two algorithms, ROC analysis is used to asses the relative value of wavelet features compared to simpler pixel features. No universally best feature is observed, and for imagery acquired under uncontrolled lighting, pixels perform slightly better than wavelets. The Cascade Classifier is found to be impossible to train in the same fashion as the other algorithms, but it is also found to perform very well using a training configuration supplied along with the algorithm as part of the OpenCV library.

Original language	English
Title of host publication	IEEE Conference on Biometrics
Subtitle of host publication	Theory, Applications and Systems, BTAS'07
DOIs	https://doi.org/10.1109/BTAS.2007.4401950
State	Published - 2007
Event	1st IEEE International Conference on Biometrics: Theory, Applications, and Systems, BTAS '07 - Crystal City, VA, United States Duration: Sep 27 2007 → Sep 29 2007

Publication series

Name	IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07

Conference

Conference	1st IEEE International Conference on Biometrics: Theory, Applications, and Systems, BTAS '07
Country/Territory	United States
City	Crystal City, VA
Period	09/27/07 → 09/29/07

Access to Document

10.1109/BTAS.2007.4401950

Cite this

@inproceedings{272b9235a6d342fdb1dbd1e12e0bf8b6,

title = "Face detection algorithm and feature performance on FRGC 2.0 imagery",

abstract = "The performance of three well known face detection algorithms and four alternative types of features are characterized using face data from the Face Recognition Grand Challenge. The three algorithms are a Semi-Naive Bayesian Classifier, a neural network called a SNoW, and a Cascade Classifier using Haar wavelets. For the first two algorithms, ROC analysis is used to asses the relative value of wavelet features compared to simpler pixel features. No universally best feature is observed, and for imagery acquired under uncontrolled lighting, pixels perform slightly better than wavelets. The Cascade Classifier is found to be impossible to train in the same fashion as the other algorithms, but it is also found to perform very well using a training configuration supplied along with the algorithm as part of the OpenCV library.",

author = "Beveridge, \{J. Ross\} and Andres Alvarez and Jilmil Saraf and Ward Fisher and Flynn, \{Patrick J.\} and James Gentile",

year = "2007",

doi = "10.1109/BTAS.2007.4401950",

language = "English",

isbn = "9781424415977",

series = "IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07",

booktitle = "IEEE Conference on Biometrics",

note = "1st IEEE International Conference on Biometrics: Theory, Applications, and Systems, BTAS '07 ; Conference date: 27-09-2007 Through 29-09-2007",

}

Beveridge, JR, Alvarez, A, Saraf, J, Fisher, W, Flynn, PJ & Gentile, J 2007, Face detection algorithm and feature performance on FRGC 2.0 imagery. in IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07., 4401950, IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07, 1st IEEE International Conference on Biometrics: Theory, Applications, and Systems, BTAS '07, Crystal City, VA, United States, 09/27/07. https://doi.org/10.1109/BTAS.2007.4401950

Face detection algorithm and feature performance on FRGC 2.0 imagery. / Beveridge, J. Ross; Alvarez, Andres; Saraf, Jilmil et al.
IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07. 2007. 4401950 (IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07).

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

TY - GEN

T1 - Face detection algorithm and feature performance on FRGC 2.0 imagery

AU - Beveridge, J. Ross

AU - Alvarez, Andres

AU - Saraf, Jilmil

AU - Fisher, Ward

AU - Flynn, Patrick J.

AU - Gentile, James

PY - 2007

Y1 - 2007

N2 - The performance of three well known face detection algorithms and four alternative types of features are characterized using face data from the Face Recognition Grand Challenge. The three algorithms are a Semi-Naive Bayesian Classifier, a neural network called a SNoW, and a Cascade Classifier using Haar wavelets. For the first two algorithms, ROC analysis is used to asses the relative value of wavelet features compared to simpler pixel features. No universally best feature is observed, and for imagery acquired under uncontrolled lighting, pixels perform slightly better than wavelets. The Cascade Classifier is found to be impossible to train in the same fashion as the other algorithms, but it is also found to perform very well using a training configuration supplied along with the algorithm as part of the OpenCV library.

AB - The performance of three well known face detection algorithms and four alternative types of features are characterized using face data from the Face Recognition Grand Challenge. The three algorithms are a Semi-Naive Bayesian Classifier, a neural network called a SNoW, and a Cascade Classifier using Haar wavelets. For the first two algorithms, ROC analysis is used to asses the relative value of wavelet features compared to simpler pixel features. No universally best feature is observed, and for imagery acquired under uncontrolled lighting, pixels perform slightly better than wavelets. The Cascade Classifier is found to be impossible to train in the same fashion as the other algorithms, but it is also found to perform very well using a training configuration supplied along with the algorithm as part of the OpenCV library.

UR - https://www.scopus.com/pages/publications/48649108892

U2 - 10.1109/BTAS.2007.4401950

DO - 10.1109/BTAS.2007.4401950

M3 - Conference contribution

AN - SCOPUS:48649108892

SN - 9781424415977

T3 - IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07

BT - IEEE Conference on Biometrics

T2 - 1st IEEE International Conference on Biometrics: Theory, Applications, and Systems, BTAS '07

Y2 - 27 September 2007 through 29 September 2007

ER -

Face detection algorithm and feature performance on FRGC 2.0 imagery

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this