Hostname: page-component-6bf8c574d5-2jptb Total loading time: 0 Render date: 2025-02-20T10:56:56.525Z Has data issue: false hasContentIssue false
  • English
  • Français

For deep networks, the whole equals the sum of the parts

Published online by Cambridge University Press:  06 December 2023

Philip J. Kellman Open the ORCID record for Philip J. Kellman [Opens in a new window] ,
Nicholas Baker ,
Patrick Garrigan ,
Austin Phillips  and
Hongjing Lu
Philip J. Kellman
Affiliation:
Department of Psychology and David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA, USA [email protected]; https://kellmanlab.psych.ucla.edu/
Nicholas Baker
Affiliation:
Department of Psychology, Loyola University of Chicago, Chicago, IL, USA [email protected]; https://www.luc.edu/psychology/people/staff/facultyandstaff/nicholasbaker/
Patrick Garrigan
Affiliation:
Department of Psychology, St. Joseph's University, Philadelphia, PA, USA [email protected]; https://sjupsych.org/faculty_pg.php
Austin Phillips
Affiliation:
Department of Psychology, University of California, Los Angeles, Los Angeles, CA, USA [email protected]; https://kellmanlab.psych.ucla.edu/
Hongjing Lu
Affiliation:
Department of Psychology and Department of Statistics, University of California, Los Angeles, Los Angeles, CA, USA [email protected]; https://cvl.psych.ucla.edu/
Get access Rights & Permissions [Opens in a new window]

Abstract

Deep convolutional networks exceed humans in sensitivity to local image properties, but unlike biological vision systems, do not discover and encode abstract relations that capture important properties of objects and events in the world. Coupling network architectures with additional machinery for encoding abstract relations will make deep networks better models of human abilities and more versatile and capable artificial devices.

Type
Open Peer Commentary
Information
Behavioral and Brain Sciences , Volume 46 , 2023 , e396
Check for updates
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Baker, N., Garrigan, P., Phillips, A., & Kellman, P. J. (2023). Configural relations in humans and deep convolutional neural networks. Frontiers in Artificial Intelligence, 5, 961595. doi:10.3389/frai.2022.961595CrossRefGoogle ScholarPubMed
Baker, N., & Kellman, P. J. (2018). Abstract shape representation in human visual perception. Journal of Experimental Psychology: General, 147(9), 1295.CrossRefGoogle ScholarPubMed
Baker, N., Lu, H., Erlikhman, G., & Kellman, P. J. (2018). Deep convolutional networks do not classify based on global object shape. PLoS Computational Biology, 14(12), e1006613.CrossRefGoogle Scholar
Baker, N., Lu, H., Erlikhman, G., & Kellman, P. J. (2020). Local features and global shape information in object classification by deep convolutional neural networks. Vision Research, 172, 46–61.CrossRefGoogle ScholarPubMed
Fan, J. E., Yamins, D. L., & Turk-Browne, N. B. (2018). Common object representations for visual production and recognition. Cognitive Science, 42(8), 2670–2698.CrossRefGoogle ScholarPubMed
Geirhos, R., Rubisch, P., Michaelis, C., Bethge, M., Wichmann, F. A., & Brendel, W. (2019) ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness. In International conference on learning representations (ICLR), https://arxiv.org/abs/1811.12231Google Scholar
Kellman, P. J., & Massey, C. M. (2013). Perceptual learning, cognition, and expertise. In Ross, B. H. (Ed.), The psychology of learning and motivation (Vol. 58, pp. 117–165). Elsevier.Google Scholar
Kubilius, J., Bracci, S., & de Beeck, H. P. O. (2016). Deep neural networks as a computational model for human shape sensitivity. PLoS Computational Biology, 12(4), e1004896.CrossRefGoogle ScholarPubMed
Malhotra, G., Dujmović, M., & Bowers, J. S. (2022). Feature blindness: A challenge for understanding and modelling visual object recognition. PLoS Computational Biology, 18(5), e1009572.CrossRefGoogle ScholarPubMed
Malhotra, G., Dujmović, M., Hummel, J., & Bowers, J. S. (2021). The contrasting shape representations that support object recognition in humans and CNNs. arXiv preprint, https://doi.org/10.1101/2021.12.14.472546Google Scholar
Puebla, G., & Bowers, J. (2021a). Can deep convolutional neural networks support relational reasoning in the same-different task? arXiv preprint, https://doi.org/10.1101/2021.09.03.458919Google Scholar
Puebla, G., & Bowers, J. (2021b). Can deep convolutional neural networks learn same-different relations?. bioRxiv, 2021-04.CrossRefGoogle Scholar
Saxe, A. M., McClelland, J. L., & Ganguli, S. (2019). A mathematical theory of semantic development in deep neural networks. Proceedings of the National Academy of Sciences, 116(23), 11537–11546.CrossRefGoogle ScholarPubMed