Contrast sensitivity functions in autoencoders

Mostra el registre complet Registre parcial de l'ítem

• dc.contributor.author Li, Qiang
• dc.contributor.author Gomez-Villa, Alex
• dc.contributor.author Bertalmío, Marcelo
• dc.contributor.author Malo, Jesús
• dc.date.accessioned 2023-03-01T07:26:04Z
• dc.date.available 2023-03-01T07:26:04Z
• dc.date.issued 2022
• dc.description.abstract Three decades ago, Atick et al. suggested that human frequency sensitivity may emerge from the enhancement required for a more efficient analysis of retinal images. Here we reassess the relevance of low-level vision tasks in the explanation of the contrast sensitivity functions (CSFs) in light of 1) the current trend of using artificial neural networks for studying vision, and 2) the current knowledge of retinal image representations. As a first contribution, we show that a very popular type of convolutional neural networks (CNNs), called autoencoders, may develop human-like CSFs in the spatiotemporal and chromatic dimensions when trained to perform some basic low-level vision tasks (like retinal noise and optical blur removal), but not others (like chromatic) adaptation or pure reconstruction after simple bottlenecks). As an illustrative example, the best CNN (in the considered set of simple architectures for enhancement of the retinal signal) reproduces the CSFs with a root mean square error of 11% of the maximum sensitivity. As a second contribution, we provide experimental evidence of the fact that, for some functional goals (at low abstraction level), deeper CNNs that are better in reaching the quantitative goal are actually worse in replicating human-like phenomena (such as the CSFs). This low-level result (for the explored networks) is not necessarily in contradiction with other works that report advantages of deeper nets in modeling higher level vision goals. However, in line with a growing body of literature, our results suggests another word of caution about CNNs in vision science because the use of simplified units or unrealistic architectures in goal optimization may be a limitation for the modeling and understanding of human vision.
• dc.description.sponsorship Partially funded by these grants from GVA/AEI/FEDER/EU: MICINN DPI2017- 89867-C2-2-R, MICINN PID2020-118071GB-I00, and GVA Grisolía-P/2019/035 (for JM and QL), and MICINN PGC2018-099651-B-I00 (for A.G.V. and M.B.).
• dc.format.mimetype application/pdf
• dc.identifier.citation Li Q, Gomez-Villa A, Bertalmío M, Malo J. Contrast sensitivity functions in autoencoders. J Vis. 2022;22(6):8. DOI: 10.1167/jov.22.6.8
• dc.identifier.doi http://dx.doi.org/10.1167/jov.22.6.8
• dc.identifier.issn 1534-7362
• dc.identifier.uri http://hdl.handle.net/10230/55982
• dc.language.iso eng
• dc.publisher Association for Research in Vision and Ophthalmology (ARVO)
• dc.relation.ispartof Journal of Vision. 2022;22(6):8.
• dc.relation.projectID info:eu-repo/grantAgreement/ES/2PE/PID2020-118071GB-I00
• dc.relation.projectID info:eu-repo/grantAgreement/ES/2PE/DPI2017-89867-C2-2-R
• dc.relation.projectID info:eu-repo/grantAgreement/ES/2PE/PGC2018-099651-B-I00
• dc.rights Copyright 2022 The Authors. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
• dc.rights.accessRights info:eu-repo/semantics/openAccess
• dc.rights.uri http://creativecommons.org/licenses/by-nc-nd/4.0/
• dc.subject.keyword spatiotemporal and chromatic contrast sensitivity
• dc.subject.keyword convolutional autoencoders
• dc.subject.keyword modulation transfer function
• dc.subject.keyword noisy cones
• dc.subject.keyword deblurring and denoising
• dc.subject.keyword chromatic adaptation
• dc.subject.keyword natural images
• dc.subject.keyword statistical goals
• dc.subject.keyword architectures
• dc.title Contrast sensitivity functions in autoencoders
• dc.type info:eu-repo/semantics/article
• dc.type.version info:eu-repo/semantics/publishedVersion