File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: Counteracting Adversarial Attacks in Autonomous Driving

TitleCounteracting Adversarial Attacks in Autonomous Driving
Authors
KeywordsAdversarial Defense
Autonomous System
Local Smoothness
Robust Stereo Vision
Issue Date2020
Citation
IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD, 2020, v. 2020-November, article no. 9256443 How to Cite?
AbstractIn this paper, we focus on studying robust deep stereo vision of autonomous driving systems and counteracting adversarial attacks against it. Autonomous system operation requires real-time processing of measurement data which often contain significant uncertainties and noise. Adversarial attacks have been widely studied to simulate these perturbations in recent years. To counteract these attacks in autonomous systems, a novel defense method is proposed in this paper. A stereo-regularizer is proposed to guide the model to learn the implicit relationship between the left and right images of the stereo-vision system. Univariate and multivariate functions are adopted to characterize the relationships between the two input images and the object detection model. The regularizer is then relaxed to its upper bound to improve adversarial robustness. Furthermore, the upper bound is approximated by the remainder of its Taylor expansion to improve the local smoothness of the loss surface. The model parameters are trained via adversarial training with the novel regularization term. Our method exploits basic knowledge from the physical world, i.e., the mutual constraints of the two images in the stereo-based system. As such, outliers can be detected and defended with high accuracy and efficiency. Numerical experiments demonstrate that the proposed method offers superior performance when compared with traditional adversarial training methods in state-of-the-art stereo-based 3D object detection models for autonomous vehicles.
Persistent Identifierhttp://hdl.handle.net/10722/336259
ISSN
2023 SCImago Journal Rankings: 0.894
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorSun, Qi-
dc.contributor.authorRao, Arjun Ashok-
dc.contributor.authorYao, Xufeng-
dc.contributor.authorYu, Bei-
dc.contributor.authorHu, Shiyan-
dc.date.accessioned2024-01-15T08:24:58Z-
dc.date.available2024-01-15T08:24:58Z-
dc.date.issued2020-
dc.identifier.citationIEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD, 2020, v. 2020-November, article no. 9256443-
dc.identifier.issn1092-3152-
dc.identifier.urihttp://hdl.handle.net/10722/336259-
dc.description.abstractIn this paper, we focus on studying robust deep stereo vision of autonomous driving systems and counteracting adversarial attacks against it. Autonomous system operation requires real-time processing of measurement data which often contain significant uncertainties and noise. Adversarial attacks have been widely studied to simulate these perturbations in recent years. To counteract these attacks in autonomous systems, a novel defense method is proposed in this paper. A stereo-regularizer is proposed to guide the model to learn the implicit relationship between the left and right images of the stereo-vision system. Univariate and multivariate functions are adopted to characterize the relationships between the two input images and the object detection model. The regularizer is then relaxed to its upper bound to improve adversarial robustness. Furthermore, the upper bound is approximated by the remainder of its Taylor expansion to improve the local smoothness of the loss surface. The model parameters are trained via adversarial training with the novel regularization term. Our method exploits basic knowledge from the physical world, i.e., the mutual constraints of the two images in the stereo-based system. As such, outliers can be detected and defended with high accuracy and efficiency. Numerical experiments demonstrate that the proposed method offers superior performance when compared with traditional adversarial training methods in state-of-the-art stereo-based 3D object detection models for autonomous vehicles.-
dc.languageeng-
dc.relation.ispartofIEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD-
dc.subjectAdversarial Defense-
dc.subjectAutonomous System-
dc.subjectLocal Smoothness-
dc.subjectRobust Stereo Vision-
dc.titleCounteracting Adversarial Attacks in Autonomous Driving-
dc.typeConference_Paper-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1145/3400302.3415758-
dc.identifier.scopuseid_2-s2.0-85097923161-
dc.identifier.volume2020-November-
dc.identifier.spagearticle no. 9256443-
dc.identifier.epagearticle no. 9256443-
dc.identifier.isiWOS:000671087100015-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats