Size and contrast have only small effects on the responses to faces of neurons in the cortex of the superior temporal sulcus of the monkey

E. T. Rolls,G. C. Baylis
DOI: https://doi.org/10.1007/BF00243828
1986-01-01
Experimental Brain Research
Abstract:There is a population of neurons in the cortex in the middle and anterior part of the superior temporal sulcus (STS) of the monkey with responses which are selective for faces. To investigate whether the responses of these neurons show some of the perceptual properties of face recognition such as tolerance to changes in the size and contrast of the face, the effects of alteration of the size and contrast of an effective face stimulus on the responses of these neurons were analysed quantitatively in macaque monkeys. First, it was shown that the majority of these neurons had responses which were relatively invariant with respect to the size of the stimulus. The median size change tolerated with a response of greater than half the maximal response was 12 times. Second, it was found that for a few of these neurons, the size of the face did affect the neuronal response. For most of these neurons, it was found that when the size of the image and its distance were altered, the neuronal response was related to the retinal angle subtended by the image. But for four neurons the absolute size of the image determined the magnitude of the neuronal response, independently of the distance of the image. Thus these four neurons showed size constancy. It is suggested that these neurons would be useful as part of a face recognition system, because only objects in a certain absolute size range should normally be classified as faces. Third, the responses of the neurons were relatively invariant with respect to the contrast of the face. The mean contrast at which the neurons still responded with more than half the maximal response was 0.26. Fourth, the responses of the neurons were relatively invariant with respect to the sign of the contrast of the face, that is the neurons responded to negative as well as to positive images of faces. Fifth, the neurons typically responded to a face when the information in it had been reduced from 3D to a 2D representation in gray on a monitor, with a response which was on average 0.5 that to a real face. These results show that the responses of these neurons have some of the invariant properties with respect to size and contrast alteration shown by face perception, and show that their processing is at a level which would be useful in face recognition. The results also show that the responses of these neurons are not simply to local contour information.
What problem does this paper attempt to address?