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

Dolhasz, Alan and Harvey, Carlo and Williams, Ian (2020) Learning to Observe: Approximating Human Perceptual Thresholds for Detection of Suprathreshold Image Transformations. In: Computer.

Text (Accepted Camera-ready paper)
09479.pdf - Accepted Version

Download (4MB)
Text (Supplementary materials (accepted version))
09479-supp.pdf - Supplemental Material

Download (20MB)


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.

Item Type: Conference or Workshop Item (Paper)
Additional Information: Data available at: https://github.com/dmt-lab/learning-to-observe
27 February 2020Accepted
27 March 2020Published Online
Subjects: CAH11 - computing > CAH11-01 - computing > CAH11-01-01 - computer science
Divisions: Faculty of Computing, Engineering and the Built Environment > School of Computing and Digital Technology
Depositing User: Alan Dolhasz
Date Deposited: 24 Apr 2020 10:52
Last Modified: 22 Mar 2023 12:01
URI: https://www.open-access.bcu.ac.uk/id/eprint/9002

Actions (login required)

View Item View Item


In this section...