On the Robustness of Convolutional Neural Networks to Internal Architecture and Weight Perturbations
Name
CBMM-Memo-065.pdf
Size
687.76 KB
Format
Adobe PDF
Checksum (MD5)
40f8d44b9364a57e10380c208854ad92
Author(s)
Cheney, Nicholas
Schrimpf, Martin
Kreiman, Gabriel
Date Issued
April 3, 2017
Publisher
Center for Brains, Minds and Machines (CBMM), arXiv
Citation
arXiv:1703.08245
Series/Report no.
CBMM Memo Series;065
Abstract
Deep convolutional neural networks are generally regarded as robust function approximators. So far, this intuition is based on perturbations to external stimuli such as the images to be classified. Here we explore the robustness of convolutional neural networks to perturbations to the internal weights and architecture of the network itself. We show that convolutional networks are surprisingly robust to a number of internal perturbations in the higher convolutional layers but the bottom convolutional layers are much more fragile. For instance, Alexnet shows less than a 30% decrease in classification performance when randomly removing over 70% of weight connections in the top convolutional or dense layers but performance is almost at chance with the same perturbation in the first convolutional layer. Finally, we suggest further investigations which could continue to inform the robustness of convolutional networks to internal perturbations.
Terms of Use
Attribution-NonCommercial-ShareAlike 3.0 United States
Persistent DSpace Link
