Does a Bayesian model of V1 contrast coding offer a neurophysiological account of human contrast discrimination?

Abstract

The dipper effect for contrast discrimination provides strong evidence that the underlying neural response is accelerating at low contrasts and saturating at high contrasts. The contrast-response functions of V1 neurons do have this sigmoidal shape, but individual neurons do not generally have a dynamic range wide enough to account for the dipper effect. This paper presents a Bayesian model of neurons in monkey V1, whose contrast-response function is described by a modified Naka-Rushton with multiplicative noise. It is shown that a model of groups of twelve or more neurons gives a reasonable explanation of the psychophysical data of two observers, but there is a large systematic error which is apparently due to the shape of the distribution of the monkey's sensitivity parameter, c50. A further model provides a better fit to the data by sacrificing strict adherence to V1 neuronal parameters and, instead using an arbitrary bimodal c50 distribution, perhaps reflecting differences between M- and P-cells.

Differential attentional modulation of cortical responses to S-cone and luminance stimuli

“…The other free parameter in this equation (n, the response exponent) was fixed at a value of 2, which is standard for population response modeling of this kind (Boynton, Demb, Glover, & Heeger, 1999;Chirimuuta & Tolhurst, 2005). In general, our data were well fit by these equations.…”

Fixed vs. variable noise in 2AFC contrast discrimination: lessons from psychometric functions

“…Contrast discrimination under the transducer model with fixed noise is assumed to occur when the increment x added to a stimulus of level x is such that μ(log[10 x + 10 x ]) − μ(x) = k, where k is the response magnitude associated with a just noticeable difference (Chen and Tyler, 2001;Chirimuuta and Tolhurst, 2005;Huang and Dobkins, 2005;Nachmias and Sansbury, 1974;Solomon et al, 1999;Wilson, 1980). The formal analysis presented by García-Pérez and Alcalá-Quintana (2007) showed that the unknown value of the response magnitude k is related to the performance level in a 2AFC contrast discrimination task, that is, to the percentage correct with respect to which discrimination is defined.…”

“…The rest of the curve for increasing x is obtained through linear interpolation across subsequent points at ordinates y i = k × i and abscissae x i = log[10 x i−1 + 10 s(x i−1 ) ] for integer i > 1, where s is the cubic spline in the corresponding upper panel. The part of the curve on the left of the starting point is impossible to determine accurately because this region is below the equivalent of a hard threshold (Chirimuuta and Tolhurst, 2005), but this region is important for determining the lower tail of psychometric functions. To fill this gap, we followed two different strategies because neither of them yielded acceptable results for all subjects: For subjects 1 and 2, we completed the curve with power functions with exponents of 5.3 and 8.5, respectively; for subject 3, we worked backwards from the starting point defining μ(x) = μ(log[10 x + 10 s(x) ]) − k for all x < T .…”

Evaluation of a multiscale color model for visual difference prediction

“…We then, must estimate how much each value of C might contribute toward the visibility of the difference between the pictures. We hypothesize that visibility depends not just on C, but that it follows Weber's Law: i.e., we evaluate each C value against the familiar "dipper function" for contrast discrimination for sinusoidal gratings [Campbell and Kulikowski 1966;Nachmias and Sansbury 1974;Tolhurst and Barfield, 1978;Legge 1981;Legge and Foley 1980;Meese, 2004;Chirimuuta and Tolhurst, 2005]. Figure 1 shows such a "dipper function".…”