A novel feature-scrambling approach reveals the capacity of convolutional neural networks to learn spatial relations

  • Convolutional neural networks (CNNs) are one of the most successful computer vision systems to solve object recognition. Furthermore, CNNs have major applications in understanding the nature of visual representations in the human brain. Yet it remains poorly understood how CNNs actually make their decisions, what the nature of their internal representations is, and how their recognition strategies differ from humans. Specifically, there is a major debate about the question of whether CNNs primarily rely on surface regularities of objects, or whether they are capable of exploiting the spatial arrangement of features, similar to humans. Here, we develop a novel feature-scrambling approach to explicitly test whether CNNs use the spatial arrangement of features (i.e. object parts) to classify objects. We combine this approach with a systematic manipulation of effective receptive field sizes of CNNs as well as minimal recognizable configurations (MIRCs) analysis. In contrast to much previous literature, we provide evidence that CNNs are in fact capable of using relatively long-range spatial relationships for object classification. Moreover, the extent to which CNNs use spatial relationships depends heavily on the dataset, e.g. texture vs. sketch. In fact, CNNs even use different strategies for different classes within heterogeneous datasets (ImageNet), suggesting CNNs have a continuous spectrum of classification strategies. Finally, we show that CNNs learn the spatial arrangement of features only up to an intermediate level of granularity, which suggests that intermediate rather than global shape features provide the optimal trade-off between sensitivity and specificity in object classification. These results provide novel insights into the nature of CNN representations and the extent to which they rely on the spatial arrangement of features for object classification.

Download full text files

Export metadata

Additional Services

Share in Twitter Search Google Scholar
Author:Amr FarahatORCiD, Felix EffenbergerORCiDGND, Martin VinckORCiD
Parent Title (English):Neural networks
Place of publication:Amsterdam [u.a.]
Document Type:Article
Date of Publication (online):2023/08/18
Date of first Publication:2023/08/18
Publishing Institution:Universit├Ątsbibliothek Johann Christian Senckenberg
Release Date:2023/09/20
Tag:CNNs; Computer vision; Object recognition; Shape representations; Texture bias; Visual cortex
Page Number:15
First Page:400
Last Page:414
Institutes:Wissenschaftliche Zentren und koordinierte Programme / Frankfurt Institute for Advanced Studies (FIAS)
Dewey Decimal Classification:1 Philosophie und Psychologie / 15 Psychologie / 150 Psychologie
Licence (German):License LogoCreative Commons - CC BY - Namensnennung 4.0 International