On the Robustness of Convolutional Neural Networks to Internal Architecture and Weight Perturbations

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.

under review at ICML 2017

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Explaining and Harnessing Adversarial Examples
• Improving neural networks by preventing co-adaptation of feature detectors
• Caffe: Convolutional Architecture for Fast Feature Embedding
• Understanding deep learning requires rethinking generalization
• An Empirical Evaluation of Deep Learning on Highway Driving

Cited by in corpus (8)

• Under the Hood of Neural Networks: Characterizing Learned Representations by Functional Neuron Populations and Network Ablations
• A Neural Spiking Approach Compared to Deep Feedforward Networks on Stepwise Pixel Erasement
• Adversarial Robustness in Deep Learning: Attacks on Fragile Neurons
• Paradox in Deep Neural Networks: Similar yet Different while Different yet Similar
• Stabilizing Inputs to Approximated Nonlinear Functions for Inference with Homomorphic Encryption in Deep Neural Networks
• How Do You Act? An Empirical Study to Understand Behavior of Deep Reinforcement Learning Agents
• Non-Singular Adversarial Robustness of Neural Networks
• Geometric algorithms for predicting resilience and recovering damage in neural networks