Size is a fundamental attribute of objects. Many neurons are very selective for the size of objects presented in their classical receptive fields (CRFs); however, the perception of object size also depends on the visual context including the size of other elements in the field and the distance of the object from the observer. We propose to investigate the mechanisms for the mediation of size perceptIon in the macular representations in cortical areas for object vision, V1, V2, V4, and IT in awake, behaving macaque monkeys (Macaca mulatta). In our preliminary studies we found 2 main classes of distance modulated neurons: nearness cells in which the response increases with proximity to the monkey, and farness cells in which the response increases with distance from the monkey. These neurons typically are tuned to the same retinal image size at different distances. We will complete our survey of the influence of distance and also search for neurons that exhibit true size constancy, i.e. cells that are tuned to absolute object size at different distances. We also plan to test for the possible influence of attention by engaging the monkeys in size comparison tasks. We will determine how the visual context influences the responses of neurons to different sized stimuli presented within the CRF. We plan 2 approaches to this problem through the Ebbinghaus illusion and the virtual laboratory. Each of these provides a way to probe the non- classical receptive field (non-CRF) to reveal its role in the mechanisms of size perception. The perceived size of objects is influenced by the size of surrounding objects in the Ebbinghaus illusion. We will test the influence of the non-CRF in size-selective cells by presenting different sized and shaped stimuli in the surround while measuring the responses to different sized stimuli presented within the CRF. In another test of the influence of the visual context, we will create a virtual laboratory in the form of a computer graphics model of our actual experimental set- up. The advantage of the computer graphics model is that stereoscopic depth and perspective can be varied as independent parameters to test the influence of these cues on the monkey's perception of size as well as their influence on the distance modulation or size constancy of neurons in V1,V2,V4 and IT. We will also test the effects of binocular disparity and perspective with wire-frame style images in which color, background shading and texture will be removed from the computer graphics display. Finally, we will map the extent of the surrounding non-CRF by masking parts of the visual contextual display while measuring the responses to different sized stimuli presented within the CRF.
