Learning to Observe: Approximating Human Perceptual Thresholds for Detection of Suprathreshold Image Transformations

TL;DR

This paper introduces a method to approximate human perceptual thresholds for detecting local image transformations, using a combination of subjective studies and transformation-equivariant learning, enabling accurate detection of exposure shifts.

Contribution

It presents a novel approach to model human perceptual thresholds for image transformations by combining subjective experiments with advanced representation learning techniques.

Findings

Achieved an average error of 0.1148 exposure stops in threshold approximation.

Successfully trained a dense classifier to detect local exposure shifts.

Model generalizes to detect various local transformations.

Abstract

Many tasks in computer vision are often calibrated and evaluated relative to human perception. In this paper, we propose to directly approximate the perceptual function performed by human observers completing a visual detection task. Specifically, we present a novel methodology for learning to detect image transformations visible to human observers through approximating perceptual thresholds. To do this, we carry out a subjective two-alternative forced-choice study to estimate perceptual thresholds of human observers detecting local exposure shifts in images. We then leverage transformation equivariant representation learning to overcome issues of limited perceptual data. This representation is then used to train a dense convolutional classifier capable of detecting local suprathreshold exposure shifts - a distortion common to image composites. In this context, our model can approximate perceptual thresholds with an average error of 0.1148 exposure stops between empirical and predicted thresholds. It can also be trained to detect a range of different local transformations.