File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1007/978-3-030-59013-0_23
- Scopus: eid_2-s2.0-85091558595
- Find via
Supplementary
-
Citations:
- Scopus: 0
- Appears in Collections:
Conference Paper: Understanding object detection through an adversarial lens
Title | Understanding object detection through an adversarial lens |
---|---|
Authors | |
Keywords | Adversarial robustness Attack evaluation framework Deep neural networks Object detection |
Issue Date | 2020 |
Citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2020, v. 12309 LNCS, p. 460-481 How to Cite? |
Abstract | Deep neural networks based object detection models have revolutionized computer vision and fueled the development of a wide range of visual recognition applications. However, recent studies have revealed that deep object detectors can be compromised under adversarial attacks, causing a victim detector to detect no object, fake objects, or mislabeled objects. With object detection being used pervasively in many security-critical applications, such as autonomous vehicles and smart cities, we argue that a holistic approach for an in-depth understanding of adversarial attacks and vulnerabilities of deep object detection systems is of utmost importance for the research community to develop robust defense mechanisms. This paper presents a framework for analyzing and evaluating vulnerabilities of the state-of-the-art object detectors under an adversarial lens, aiming to analyze and demystify the attack strategies, adverse effects, and costs, as well as the cross-model and cross-resolution transferability of attacks. Using a set of quantitative metrics, extensive experiments are performed on six representative deep object detectors from three popular families (YOLOv3, SSD, and Faster R-CNN) with two benchmark datasets (PASCAL VOC and MS COCO). We demonstrate that the proposed framework can serve as a methodical benchmark for analyzing adversarial behaviors and risks in real-time object detection systems. We conjecture that this framework can also serve as a tool to assess the security risks and the adversarial robustness of deep object detectors to be deployed in real-world applications. |
Persistent Identifier | http://hdl.handle.net/10722/343318 |
ISSN | 2023 SCImago Journal Rankings: 0.606 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Chow, Ka Ho | - |
dc.contributor.author | Liu, Ling | - |
dc.contributor.author | Gursoy, Mehmet Emre | - |
dc.contributor.author | Truex, Stacey | - |
dc.contributor.author | Wei, Wenqi | - |
dc.contributor.author | Wu, Yanzhao | - |
dc.date.accessioned | 2024-05-10T09:07:09Z | - |
dc.date.available | 2024-05-10T09:07:09Z | - |
dc.date.issued | 2020 | - |
dc.identifier.citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2020, v. 12309 LNCS, p. 460-481 | - |
dc.identifier.issn | 0302-9743 | - |
dc.identifier.uri | http://hdl.handle.net/10722/343318 | - |
dc.description.abstract | Deep neural networks based object detection models have revolutionized computer vision and fueled the development of a wide range of visual recognition applications. However, recent studies have revealed that deep object detectors can be compromised under adversarial attacks, causing a victim detector to detect no object, fake objects, or mislabeled objects. With object detection being used pervasively in many security-critical applications, such as autonomous vehicles and smart cities, we argue that a holistic approach for an in-depth understanding of adversarial attacks and vulnerabilities of deep object detection systems is of utmost importance for the research community to develop robust defense mechanisms. This paper presents a framework for analyzing and evaluating vulnerabilities of the state-of-the-art object detectors under an adversarial lens, aiming to analyze and demystify the attack strategies, adverse effects, and costs, as well as the cross-model and cross-resolution transferability of attacks. Using a set of quantitative metrics, extensive experiments are performed on six representative deep object detectors from three popular families (YOLOv3, SSD, and Faster R-CNN) with two benchmark datasets (PASCAL VOC and MS COCO). We demonstrate that the proposed framework can serve as a methodical benchmark for analyzing adversarial behaviors and risks in real-time object detection systems. We conjecture that this framework can also serve as a tool to assess the security risks and the adversarial robustness of deep object detectors to be deployed in real-world applications. | - |
dc.language | eng | - |
dc.relation.ispartof | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | - |
dc.subject | Adversarial robustness | - |
dc.subject | Attack evaluation framework | - |
dc.subject | Deep neural networks | - |
dc.subject | Object detection | - |
dc.title | Understanding object detection through an adversarial lens | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1007/978-3-030-59013-0_23 | - |
dc.identifier.scopus | eid_2-s2.0-85091558595 | - |
dc.identifier.volume | 12309 LNCS | - |
dc.identifier.spage | 460 | - |
dc.identifier.epage | 481 | - |
dc.identifier.eissn | 1611-3349 | - |