Abstract
The recent works on automated vehicle make and model recognition (VMMR) have embraced the use of advanced deep learning models such as convolutional neural networks. In this work, we introduce an adversarial attack against such VMMR systems through adversarially learnt patches. We demonstrate the effectiveness of the adversarial patches against VMMR through experimental evaluations on a real-world surveillance dataset. The developed adversarial patches achieve reductions of upto 37% in VMMR recall scores. It is hoped that this work shall motivate future studies in developing VMMR systems that are robust to adversarial learning-based attacks.
| Original language | English |
|---|---|
| Title of host publication | Q2SWinet 2020 - Proceedings of the 16th ACM Symposium on QoS and Security for Wireless and Mobile Networks |
| Publisher | Association for Computing Machinery, Inc |
| Pages | 115-121 |
| Number of pages | 7 |
| ISBN (Electronic) | 9781450381208 |
| DOIs | |
| State | Published - 16 Nov 2020 |
| Externally published | Yes |
Publication series
| Name | Q2SWinet 2020 - Proceedings of the 16th ACM Symposium on QoS and Security for Wireless and Mobile Networks |
|---|
Bibliographical note
Publisher Copyright:© 2020 ACM.
Keywords
- adversarial attacks
- adversarial patches
- adversarial robustness
- cyber-physical systems security
- smart city surveillance
- vehicle make and model recognition
ASJC Scopus subject areas
- Computer Networks and Communications
- Software
- Electrical and Electronic Engineering
- Safety, Risk, Reliability and Quality
