Efficient coding of natural scene statistics predicts discrimination thresholds for grayscale textures

TL;DR

This study demonstrates that an efficient coding model based on natural scene statistics accurately predicts human visual sensitivity to a wide range of grayscale texture correlations, revealing the limits of this predictive approach.

Contribution

The paper extends the efficient coding hypothesis to a high-dimensional space of grayscale correlations, providing detailed predictions and empirical validation for visual discrimination thresholds.

Findings

Predicted thresholds closely match psychophysical data (median fractional error <0.13)

Two-point correlations are most salient in natural scenes

Correlations beyond second order are not salient

Abstract

Previously, in (Hermundstad et al., 2014), we showed that when sampling is limiting, the efficient coding principle leads to a "variance is salience" hypothesis, and that this hypothesis accounts for visual sensitivity to binary image statistics. Here, using extensive new psychophysical data and image analysis, we show that this hypothesis accounts for visual sensitivity to a large set of grayscale image statistics at a striking level of detail, and also identify the limits of the prediction. We define a 66-dimensional space of local grayscale light-intensity correlations, and measure the relevance of each direction to natural scenes. The "variance is salience" hypothesis predicts that two-point correlations are most salient, and predicts their relative salience. We tested these predictions in a texture-segregation task using un-natural, synthetic textures. As predicted, correlations beyond second order are not salient, and predicted thresholds for over 300 second-order correlations match psychophysical thresholds closely (median fractional error <0.13).