Visual stream connectivity predicts assessments of image quality

TL;DR

This paper introduces a novel geometric model of early visual connectivity that predicts human image similarity judgments more accurately than existing measures, combining psychophysics, regression, and neural hypotheses.

Contribution

It formalizes the psychophysics of similarity using differential geometry and enhances predictions through regression and neural connectivity hypotheses, outperforming standard measures.

Findings

The geometric model accurately predicts human similarity judgments.

Regression on behavioral data improves prediction accuracy.

Connectivity-based models outperform traditional image fidelity measures.

Abstract

Some biological mechanisms of early vision are comparatively well understood, but they have yet to be evaluated for their ability to accurately predict and explain human judgments of image similarity. From well-studied simple connectivity patterns in early vision, we derive a novel formalization of the psychophysics of similarity, showing the differential geometry that provides accurate and explanatory accounts of perceptual similarity judgments. These predictions then are further improved via simple regression on human behavioral reports, which in turn are used to construct more elaborate hypothesized neural connectivity patterns. Both approaches outperform standard successful measures of perceived image fidelity from the literature, as well as providing explanatory principles of similarity perception.