Spectral receptive fields do not explain tuning for boundary curvature in V4

Abstract

The midlevel visual cortical area V4 in the primate is thought to be critical for the neural representation of visual shape. Several studies agree that V4 neurons respond to contour features, e.g., convexities and concavities along a shape boundary, that are more complex than the oriented segments encoded by neurons in the primary visual cortex. Here we compare two distinct approaches to modeling V4 shape selectivity: one based on a spectral receptive field (SRF) map in the orientation and spatial frequency domain and the other based on a map in an object-centered angular position and contour curvature space. We test the ability of these two characterizations to account for the responses of V4 neurons to a set of parametrically designed two-dimensional shapes recorded previously in the awake macaque. We report two lines of evidence suggesting that the SRF model does not capture the contour sensitivity of V4 neurons. First, the SRF model discards spatial phase information, which is inconsistent with the neuronal data. Second, the amount of variance explained by the SRF model was significantly less than that explained by the contour curvature model. Notably, cells best fit by the curvature model were poorly fit by the SRF model, the latter being appropriate for a subset of V4 neurons that appear to be orientation tuned. These limitations of the SRF model suggest that a full understanding of midlevel shape representation requires more complicated models that preserve phase information and perhaps deal with object segmentation.

Linear and non-linear properties of feature selectivity in V4 neurons

“…In this study we used neurophysiological responses to partially masked natural scene stimuli to explore the origins of feature selectivity in spectrally complex natural scene stimuli in area V4 of the primate. Our results indicate feature selectivity in V4 is highly non-linear, and while the quasi-linear SRF model can predict a substantial fraction of V4 response variance under some conditions, it is not sufficient to fully model selectivity for spectrally complex stimuli (Oleskiw et al, 2014 ). Importantly, the pattern of differences between the SRF model and spike-triggered masks revealed at least four key failures of the linear model, most significantly tuned suppression and a complex form of translation invariance that appears to make feature selectivity in some V4 neurons robust to fixational eye movements.…”

“…This universality is notably absent in area V4, where SRFs estimated from responses to narrowband stimuli generally fail to predict responses to broadband or natural stimuli (David et al, 2006 ; Oleskiw et al, 2014 ), a strong indicator that non-linear mechanisms substantially influence responses. Many studies have used carefully designed broadband, but not necessarily natural, stimuli to demonstrate that V4 encodes a substantial amount of information about complex 2D image properties.…”

Joint coding of shape and blur in area V4

“…To test the hypothesis that blur selectivity is an epiphenomenon of shape selectivity we first described each neuron’s shape preference in terms of tuning for boundary curvature. As done elsewhere 25 – 27 , we identified the 2D Gaussian function in a shape space spanned by angular position and curvature (APC) that best predicts responses to the preliminary shape screen conducted using sharp stimuli (see Methods: ‘Visual stimulation’). Bootstrapping was used to calculate the normalized mean-squared prediction error (Training NRMSE, Fig.…”

“…To rigorously test this hypothesis, we evaluate whether a joint model of shape and blur performs significantly better at predicting V4 responses than a marginal model that is tuned for shape alone. Importantly, this analysis builds on the well-studied APC model 22 , 25 – 27 that is known to capture tuning for boundary conformation in shape-selective V4 neurons (see Methods: ‘Analysis and model fitting’). For each cell we first fit a standard APC model to blurred responses under the assumption that neurons are invariant to boundary blur of shape stimuli.…”

Hue tuning curves in V4 change with visual context

“…The natural scene approach popularized in V1 requires approximating the response with a linear or second-order function of the image, but this is cannot be done accurately for V4. The nonlinearity of the V4 response is evidenced by the nonlinearity of the models that best predict V4 activity [5,15,16], as well as the fact that receptive fields (RFs) estimated from simple stimuli fail to predict much of the response to more complicated or natural stimuli [17][18][19]. Without access to interpretable receptive fields estimated with natural stimuli, it has not been possible to verify that experiments with artificial stimuli sets describe how V4 responds to natural images.…”