Article

The role of capacity constraints in Convolutional Neural Networks for learning random versus natural data

2022-04-01

Abstract excerpt

Convolutional neural networks (CNNs) are often described as promising models of human vision, yet they show many differences from human abilities. We focus on a superhuman capacity of top-performing CNNs, namely, their ability to learn very large datasets of random patterns. We verify that human learning on such tasks is extremely limited, even with few stimuli. We argue that the performance difference is due to C...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 23cf40b6-4147-5cff-840f-352ea1dcf34f
DOI: 10.1101/2022.03.31.486580

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

The role of capacity constraints in Convolutional Neural Networks for learning random versus natural dataDOI 10.1101/2022.03.31.486580Read

Biological convolutions improve DNN robustness to noise and generalisationSemantic proximity Read

Deep networks may capture biological behaviour for shallow, but not deep, empirical characterizationsSemantic proximity Read

Deep Problems with Neural Network Models of Human VisionSemantic proximity Read

Simulating a Primary Visual Cortex at the Front of CNNs Improves Robustness to Image PerturbationsSemantic proximity Read

Feature blindness: a challenge for understanding and modelling visual object recognitionSemantic proximity Read

Large-scale hyperparameter search for predicting human brain responses in the Algonauts challengeSemantic proximity Read

Synthesizing images to map neural networks to the human brainSemantic proximity Read

Understanding transformation tolerant visual object representations in the human brain and convolutional neural networksSemantic proximity Read

A controller-peripheral architecture and costly energy principle for learningSemantic proximity Read

A Feedforward Convolutional Neural Network Learns from Images to Covertly Attend to Contextual Cues like Human and Bayesian Ideal ObserversSemantic proximity Read

Comparing Brain-Score and ImageNet performance with responses to the scintillating grid illusionSemantic proximity Read

Model metamers illuminate divergences between biological and artificial neural networksSemantic proximity Read