Geometric robustness of deep networks: analysis and improvement

Abstract

Deep convolutional neural networks have been shown to be vulnerable to arbitrary geometric transformations. However, there is no systematic method to measure the invariance properties of deep networks to such transformations. We propose ManiFool as a simple yet scalable algorithm to measure the invariance of deep networks. In particular, our algorithm measures the robustness of deep networks to geometric transformations in a worst-case regime as they can be problematic for sensitive applications. Our extensive experimental results show that ManiFool can be used to measure the invariance of fairly complex networks on high dimensional datasets and these values can be used for analyzing the reasons for it. Furthermore, we build on Manifool to propose a new adversarial training scheme and we show its effectiveness on improving the invariance properties of deep neural networks.

References in corpus (9)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Explaining and Harnessing Adversarial Examples
• Intriguing properties of neural networks
• Deep Residual Learning for Image Recognition
• Robust Physical-World Attacks on Deep Learning Models
• Deformable Convolutional Networks
• Adversarial Transformation Networks: Learning to Generate Adversarial Examples
• DeepFool: a simple and accurate method to fool deep neural networks
• Adversarial Manipulation of Deep Representations

Cited by in corpus (5)

• Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey
• Divide, Denoise, and Defend against Adversarial Attacks
• Attacks on State-of-the-Art Face Recognition using Attentional Adversarial Attack Generative Network
• Robustness of Rotation-Equivariant Networks to Adversarial Perturbations
• Perception-in-the-Loop Adversarial Examples