Résumé
Adversarial robustness has become a topic of growing interest in machine learning since it was observed that neural networks tend to be brittle. We propose an information-geometric formulation of adversarial defense and introduce Fire, a new Fisher-Rao regularization for the categorical cross-entropy loss, which is based on the geodesic distance between the softmax outputs corresponding to natural and perturbed input features. Based on the information-geometric properties of the class of softmax distributions, we derive an explicit characterization of the Fisher-Rao Distance (FRD) for the binary and multiclass cases, and draw some interesting properties as well as connections with standard regularization metrics. Furthermore, we verify on a simple linear and Gaussian model, that all Pareto-optimal points in the accuracy-robustness region can be reached by Fire while other state-of-the-art methods fail. Empirically, we evaluate the performance of various classifiers trained with the proposed loss on standard datasets, showing up to a simultaneous 1% of improvement in terms of clean and robust performances while reducing the training time by 20% over the best-performing methods.
| Langue originale | Anglais |
|---|---|
| Pages (de - à) | 2698-2710 |
| Nombre de pages | 13 |
| Revue | IEEE Transactions on Pattern Analysis and Machine Intelligence |
| Volume | 45 |
| Numéro de publication | 3 |
| Les DOIs | |
| État | Publié - 1 mars 2023 |
Classification disciplinaire ASJC Scopus
- Logiciel
- Vision par ordinateur et reconnaissance de formes
- Mathématiques et théorie de l'informatique
- Mathématiques appliquées
- Intelligence artificielle
Empreinte digitale
Voici les principaux termes ou expressions associés à « Adversarial Robustness Via Fisher-Rao Regularization ». Ces libellés thématiques sont générés à partir du titre et du résumé de la publication. Ensemble, ils forment une empreinte digitale unique.Citer cette ressource
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver