Machine learning applied to perception

Decision-images for gender classification

Felix A. Wichmann, Arnulf B A Graf, Eero P. Simoncelli, Heinrich Bulthoff, Bernhard Schölkopf

Research output: Chapter in Book/Report/Conference proceedingConference contribution

10 Citations (Scopus)

Abstract

We study gender discrimination of human faces using a combination of psychophysical classification and discrimination experiments together with methods from machine learning. We reduce the dimensionality of a set of face images using principal component analysis, and then train a set of linear classifiers on this reduced representation (linear support vector machines (SVMs), relevance vector machines (RVMs), Fisher linear discriminant (FLD), and prototype (prot) classifiers) using human classification data. Because we combine a linear preprocessor with linear classifiers, the entire system acts as a linear classifier, allowing us to visualise the decision-image corresponding to the normal vector of the separating hyperplanes (SH) of each classifier. We predict that the female-tomaleness transition along the normal vector for classifiers closely mimicking human classification (SVM and RVM [1]) should be faster than the transition along any other direction. A psychophysical discrimination experiment using the decision images as stimuli is consistent with this prediction.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
ISBN (Print)0262195348, 9780262195348
Publication statusPublished - 2005 Jan 1
Externally publishedYes
Event18th Annual Conference on Neural Information Processing Systems, NIPS 2004 - Vancouver, BC, Canada
Duration: 2004 Dec 132004 Dec 16

Other

Other18th Annual Conference on Neural Information Processing Systems, NIPS 2004
CountryCanada
CityVancouver, BC
Period04/12/1304/12/16

Fingerprint

Learning systems
Classifiers
Support vector machines
Principal component analysis
Experiments

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Wichmann, F. A., Graf, A. B. A., Simoncelli, E. P., Bulthoff, H., & Schölkopf, B. (2005). Machine learning applied to perception: Decision-images for gender classification. In Advances in Neural Information Processing Systems Neural information processing systems foundation.

Machine learning applied to perception : Decision-images for gender classification. / Wichmann, Felix A.; Graf, Arnulf B A; Simoncelli, Eero P.; Bulthoff, Heinrich; Schölkopf, Bernhard.

Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2005.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Wichmann, FA, Graf, ABA, Simoncelli, EP, Bulthoff, H & Schölkopf, B 2005, Machine learning applied to perception: Decision-images for gender classification. in Advances in Neural Information Processing Systems. Neural information processing systems foundation, 18th Annual Conference on Neural Information Processing Systems, NIPS 2004, Vancouver, BC, Canada, 04/12/13.
Wichmann FA, Graf ABA, Simoncelli EP, Bulthoff H, Schölkopf B. Machine learning applied to perception: Decision-images for gender classification. In Advances in Neural Information Processing Systems. Neural information processing systems foundation. 2005
Wichmann, Felix A. ; Graf, Arnulf B A ; Simoncelli, Eero P. ; Bulthoff, Heinrich ; Schölkopf, Bernhard. / Machine learning applied to perception : Decision-images for gender classification. Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2005.
@inproceedings{5ae672974d844cfa96eded7e6ece075c,
title = "Machine learning applied to perception: Decision-images for gender classification",
abstract = "We study gender discrimination of human faces using a combination of psychophysical classification and discrimination experiments together with methods from machine learning. We reduce the dimensionality of a set of face images using principal component analysis, and then train a set of linear classifiers on this reduced representation (linear support vector machines (SVMs), relevance vector machines (RVMs), Fisher linear discriminant (FLD), and prototype (prot) classifiers) using human classification data. Because we combine a linear preprocessor with linear classifiers, the entire system acts as a linear classifier, allowing us to visualise the decision-image corresponding to the normal vector of the separating hyperplanes (SH) of each classifier. We predict that the female-tomaleness transition along the normal vector for classifiers closely mimicking human classification (SVM and RVM [1]) should be faster than the transition along any other direction. A psychophysical discrimination experiment using the decision images as stimuli is consistent with this prediction.",
author = "Wichmann, {Felix A.} and Graf, {Arnulf B A} and Simoncelli, {Eero P.} and Heinrich Bulthoff and Bernhard Sch{\"o}lkopf",
year = "2005",
month = "1",
day = "1",
language = "English",
isbn = "0262195348",
booktitle = "Advances in Neural Information Processing Systems",
publisher = "Neural information processing systems foundation",

}

TY - GEN

T1 - Machine learning applied to perception

T2 - Decision-images for gender classification

AU - Wichmann, Felix A.

AU - Graf, Arnulf B A

AU - Simoncelli, Eero P.

AU - Bulthoff, Heinrich

AU - Schölkopf, Bernhard

PY - 2005/1/1

Y1 - 2005/1/1

N2 - We study gender discrimination of human faces using a combination of psychophysical classification and discrimination experiments together with methods from machine learning. We reduce the dimensionality of a set of face images using principal component analysis, and then train a set of linear classifiers on this reduced representation (linear support vector machines (SVMs), relevance vector machines (RVMs), Fisher linear discriminant (FLD), and prototype (prot) classifiers) using human classification data. Because we combine a linear preprocessor with linear classifiers, the entire system acts as a linear classifier, allowing us to visualise the decision-image corresponding to the normal vector of the separating hyperplanes (SH) of each classifier. We predict that the female-tomaleness transition along the normal vector for classifiers closely mimicking human classification (SVM and RVM [1]) should be faster than the transition along any other direction. A psychophysical discrimination experiment using the decision images as stimuli is consistent with this prediction.

AB - We study gender discrimination of human faces using a combination of psychophysical classification and discrimination experiments together with methods from machine learning. We reduce the dimensionality of a set of face images using principal component analysis, and then train a set of linear classifiers on this reduced representation (linear support vector machines (SVMs), relevance vector machines (RVMs), Fisher linear discriminant (FLD), and prototype (prot) classifiers) using human classification data. Because we combine a linear preprocessor with linear classifiers, the entire system acts as a linear classifier, allowing us to visualise the decision-image corresponding to the normal vector of the separating hyperplanes (SH) of each classifier. We predict that the female-tomaleness transition along the normal vector for classifiers closely mimicking human classification (SVM and RVM [1]) should be faster than the transition along any other direction. A psychophysical discrimination experiment using the decision images as stimuli is consistent with this prediction.

UR - http://www.scopus.com/inward/record.url?scp=84898959863&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898959863&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0262195348

SN - 9780262195348

BT - Advances in Neural Information Processing Systems

PB - Neural information processing systems foundation

ER -