Demo: Automated Tracking System For LiDAR Spoofing Attacks On Moving Targets

Cao, Yulong; Ma, Jiaxiang; Fu, Kevin; Rampazzi, Sara; Mao, Morley

doi:10.14722/autosec.2021.23019

Cited by 3 publications

(3 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…LiDAR spoofing attacks. A number of works recently proved the feasibility of LiDAR spoofing attacks [2,4,6,[9][10][11]. Petit et al [9] first introduced the LiDAR sensor attack by performing relay attacks to spoof objects further than the location of the spoofer.…”

Section: Background and Related Workmentioning

confidence: 99%

“…Unfortunately, recent work has demonstrated the feasibility of LiDAR sensor spoofing attacks and their repercussions on the visual understanding of AV's [2,4,6,[9][10][11]. Alarmingly, such attacks have been shown to be successful at the model-level by managing to fool 3D object detectors to erroneously output bounding boxes for non-present obstacles (ghost objects) [4,11].…”

Section: Introductionmentioning

confidence: 99%

“…Cao et al [2] have demonstrated that it is possible to track a moving AV and perform LiDAR point spoofing -thus, resulting in attackers being able to perform temporal attacks. It remains unclear if the current defenses are able to defend against such temporal attacks which demonstrates a need for complementary approaches.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Temporal Consistency Checks to Detect LiDAR Spoofing Attacks on Autonomous Vehicle Perception

You

Hau

Demetriou

2021

Proceedings of the 1st Workshop on Security and Privacy for Mobile AI

View full text Add to dashboard Cite

LiDAR sensors are used widely in Autonomous Vehicles for better perceiving the environment which enables safer driving decisions. Recent work has demonstrated serious LiDAR spoofing attacks with alarming consequences. In particular, model-level LiDAR spoofing attacks aim to inject fake depth measurements to elicit ghost objects that are erroneously detected by 3D Object Detectors, resulting in hazardous driving decisions. In this work, we explore the use of motion as a physical invariant of genuine objects for detecting such attacks. Based on this, we propose a general methodology, 3D Temporal Consistency Check (3D-TC2), which leverages spatiotemporal information from motion prediction to verify objects detected by 3D Object Detectors. Our preliminary design and implementation of a 3D-TC2 prototype demonstrates very promising performance, providing more than 98% attack detection rate with a recall of 91% for detecting spoofed Vehicle (Car) objects, and is able to achieve real-time detection at 41Hz. CCS CONCEPTS• Security and privacy → Intrusion/anomaly detection and malware mitigation; • Computing methodologies → Computer vision; Scene anomaly detection.

show abstract