Abstract | ||
---|---|---|
We study gender discrimination of human faces using a combination of psychophysical classification and discrimination exper iments 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 (li near support vec- tor machines (SVMs), relevance vector machines (RVMs), Fisher linear discriminant (FLD), and prototype (prot) classifiers) usin g human clas- sification data. Because we combine a linear preprocessor wi th linear classifiers, the entire system acts as a linear classifier, al lowing us to visu- alise the decision-imagecorresponding to the normal vector of the separ- ating hyperplanes (SH) of each classifier. We predict that th e female-to- maleness transition along the normal vector for classifiers closely mim- icking human classification (SVM and RVM (1)) should be faste r than the transition along any other direction. A psychophysical discrimina- tion experiment using the decision images as stimuli is consistent with this prediction. |
Year | Venue | Keywords |
---|---|---|
2004 | NIPS | relevance vector machine,principal component analysis,machine learning |
Field | DocType | Citations |
Structured support vector machine,Pattern recognition,Random subspace method,Computer science,Support vector machine,Artificial intelligence,Linear discriminant analysis,Relevance vector machine,Margin classifier,Linear classifier,Machine learning,Quadratic classifier | Conference | 3 |
PageRank | References | Authors |
0.80 | 5 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
F A Wichmann | 1 | 231 | 17.54 |
Arnulf Graf | 2 | 145 | 9.85 |
Eero P Simoncelli | 3 | 1485 | 168.07 |
Heinrich H. Bülthoff | 4 | 2524 | 384.40 |
Bernhard Schölkopf | 5 | 23120 | 3091.82 |