arXiv Is Hiring a DevOps Engineer
Work on one of the world's most important websites and make an impact on open science.
View Jobs
We gratefully acknowledge support from the Simons Foundation,
member institutions
, and all contributors.
[Submitted on 16 Feb 2022 (
v1
), last revised 17 May 2022 (this version, v2)]
Title:
A Developmentally-Inspired Examination of Shape versus Texture Bias in Machines
View a PDF of the paper titled A Developmentally-Inspired Examination of Shape versus Texture Bias in Machines, by Alexa R. Tartaglini and 2 other authors
View PDF
Abstract:
Early in development, children learn to extend novel category labels to objects with the same shape, a phenomenon known as the shape bias. Inspired by these findings, Geirhos et al. (2019) examined whether deep neural networks show a shape or texture bias by constructing images with conflicting shape and texture cues. They found that convolutional neural networks strongly preferred to classify familiar objects based on texture as opposed to shape, suggesting a texture bias. However, there are a number of differences between how the networks were tested in this study versus how children are typically tested. In this work, we re-examine the inductive biases of neural networks by adapting the stimuli and procedure from Geirhos et al. (2019) to more closely follow the developmental paradigm and test on a wide range of pre-trained neural networks. Across three experiments, we find that deep neural networks exhibit a preference for shape rather than texture when tested under conditions that more closely replicate the developmental procedure.
Submission history
From: Alexa Tartaglini [
view email
]
[v1]
Wed, 16 Feb 2022 21:15:40 UTC (4,276 KB)
Tue, 17 May 2022 17:56:00 UTC (4,277 KB)
View a PDF of the paper titled A Developmentally-Inspired Examination of Shape versus Texture Bias in Machines, by Alexa R. Tartaglini and 2 other authors
View PDF
TeX Source
Other Formats
view license
Current browse context:
cs.CV
recent
|
2022-02
Change to browse by:
cs.LG
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community?
Learn more about arXivLabs
.
