Sökning: onr:"swepub:oai:DiVA.org:liu-168130" >
A Self-supervised A...
A Self-supervised Approach for Adversarial Robustness
-
- Naseer, M. (författare)
- Inception Institute of Artificial Intelligence, UAE; Data61-CSIRO, Australia; Australian National University, Australia
-
- Khan, S. (författare)
- nception Institute of Artificial Intelligence, UAE
-
- Hayat, M. (författare)
- Inception Institute of Artificial Intelligence, UAE
-
visa fler...
-
- Khan, Fahad Shahbaz, 1983- (författare)
- Linköpings universitet,Datorseende,Tekniska fakulteten,Inception Institute of Artificial Intelligence, UAE
-
- Porikli, F. (författare)
- Australian National University, Australia
-
visa färre...
-
(creator_code:org_t)
- IEEE, 2020
- 2020
- Engelska.
-
Ingår i: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). - : IEEE. - 9781728171685 ; , s. 259-268
- Relaterad länk:
-
https://urn.kb.se/re...
-
visa fler...
-
https://doi.org/10.1...
-
visa färre...
Abstract
Ämnesord
Stäng
- Adversarial examples can cause catastrophic mistakes in Deep Neural Network (DNNs) based vision systems e.g., for classification, segmentation and object detection. The vulnerability of DNNs against such attacks can prove a major roadblock towards their real-world deployment. Transferability of adversarial examples demand generalizable defenses that can provide cross-task protection. Adversarial training that enhances robustness by modifying target model’s parameters lacks such generalizability. On the other hand, different input processing based defenses fall short in the face of continuously evolving attacks. In this paper, we take the first step to combine the benefits of both approaches and propose a self-supervised adversarial training mechanism in the input space. By design, our defense is a generalizable approach and provides significant robustness against the unseen adversarial attacks (\eg by reducing the success rate of translation-invariant ensemble attack from 82.6% to 31.9% in comparison to previous state-of-the-art). It can be deployed as a plug-and-play solution to protect a variety of vision systems, as we demonstrate for the case of classification, segmentation and detection.
Ämnesord
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datorseende och robotik (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Vision and Robotics (hsv//eng)
Nyckelord
- Perturbation methods;Task analysis;Distortion;Training;Robustness;Feature extraction;Neural networks
Publikations- och innehållstyp
- ref (ämneskategori)
- kon (ämneskategori)
Hitta via bibliotek
Till lärosätets databas