Robustness of Rotation-Equivariant Networks to Adversarial Perturbations
By Beranger Dumont, Simona Maggio, Pablo Montalvo. 2018
26 citations recorded by the source corpus.
Deep neural networks have been shown to be vulnerable to adversarial examples: very small perturbations of the input having a dramatic impact on the predictions. A wealth of adversarial attacks and distance metrics to quantify the similarity between natural and adversarial images have been proposed, recently enlarging the scope of adversarial examples with geometric transformations beyond pixel-wise attacks. In this context, we investigate the robustness to adversarial attacks of new Convolutional Neural Network architectures providing equivariance to rotations. We found that rotation-equivariant networks are significantly less vulnerable to geometric-based attacks than regular networks on the MNIST, CIFAR-10, and ImageNet datasets.
Subfields: Artificial Intelligence, Molecular Biology, Safety, Robustness & Interpretability
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Max Welling
- Taco Cohen
- Christina Heinze-Deml
- Fanny Yang
- Bastiaan S. Veeling
- Erik J. Bekkers
- Zuowen Wang
- Jim Winkens
In Molecular Biology
- Christina Heinze-Deml
- Erik Poll
- Joost Visser
- Max Welling
- Taco Cohen
- A. Serban
- Jure Leskovec
- William L. Hamilton
In Safety, Robustness & Interpretability
- Fanny Yang
- Zuowen Wang
- Christina Heinze-Deml
- Erik Poll
- Joost Visser
- Max Welling
- Taco Cohen
- A. Serban
Citations in the corpus
- Rotation Equivariant CNNs for Digital Pathology (2018, methodology): “These methods show strong generalization under limited dataset size and are more robust under adversarial perturbations in rotation, translation and local geometric distortions [8].”
- Invariance-inducing regularization using worst-case transformations suffices to boost accuracy and spatial robustness (2019, background): “g.[11, 13, 34, 20, 14, 2, 10]) that worst-case prediction performance drops dramatically for neural network classifiers obtained using standard training, even for rather small transformation sets.”
- Adversarial Examples on Object Recognition (2020, background): “[40] investigated the resistance to adversarial attacks of three rotation-equivariant network architectures [32].”