Divergences in color perception between deep neural networks and humans

Ethan O Nadler; Elise Darragh-Ford; Bhargav Srinivasa Desikan; Christian Conaway; Mark Chu; Tasker Hull; Douglas Guilbeault

doi:10.1016/j.cognition.2023.105621

Divergences in color perception between deep neural networks and humans

Cognition. 2023 Dec:241:105621. doi: 10.1016/j.cognition.2023.105621. Epub 2023 Sep 14.

Authors

Ethan O Nadler¹, Elise Darragh-Ford², Bhargav Srinivasa Desikan³, Christian Conaway⁴, Mark Chu⁵, Tasker Hull⁶, Douglas Guilbeault⁷

Affiliations

¹ Carnegie Observatories, USA; Department of Physics, University of Southern California, USA. Electronic address: enadler@carnegiescience.edu.
² Kavli Institute for Particle Astrophysics and Cosmology and Department of Physics, Stanford University, USA.
³ School of Computer and Communication Sciences, École Polytechnique Fédérale de Lausanne, Switzerland; Knowledge Lab, University of Chicago, USA.
⁴ University of California, San Diego, USA.
⁵ School of the Arts, Columbia University, USA.
⁶ Psiphon Inc., Toronto, Canada.
⁷ Haas School of Business, University of California, Berkeley, USA. Electronic address: douglas.guilbeault@haas.berkeley.edu.

PMID: 37716312
DOI: 10.1016/j.cognition.2023.105621

Abstract

Deep neural networks (DNNs) are increasingly proposed as models of human vision, bolstered by their impressive performance on image classification and object recognition tasks. Yet, the extent to which DNNs capture fundamental aspects of human vision such as color perception remains unclear. Here, we develop novel experiments for evaluating the perceptual coherence of color embeddings in DNNs, and we assess how well these algorithms predict human color similarity judgments collected via an online survey. We find that state-of-the-art DNN architectures - including convolutional neural networks and vision transformers - provide color similarity judgments that strikingly diverge from human color judgments of (i) images with controlled color properties, (ii) images generated from online searches, and (iii) real-world images from the canonical CIFAR-10 dataset. We compare DNN performance against an interpretable and cognitively plausible model of color perception based on wavelet decomposition, inspired by foundational theories in computational neuroscience. While one deep learning model - a convolutional DNN trained on a style transfer task - captures some aspects of human color perception, our wavelet algorithm provides more coherent color embeddings that better predict human color judgments compared to all DNNs we examine. These results hold when altering the high-level visual task used to train similar DNN architectures (e.g., image classification versus image segmentation), as well as when examining the color embeddings of different layers in a given DNN architecture. These findings break new ground in the effort to analyze the perceptual representations of machine learning algorithms and to improve their ability to serve as cognitively plausible models of human vision. Implications for machine learning, human perception, and embodied cognition are discussed.

Keywords: Color perception; Computer vision; Deep learning; Embodied cognition; Wavelet decomposition.