Multi-Task Consistency-based Detection of Adversarial Attacks

2026-08-07

AI Summary

This work addresses the vulnerability of deep neural networks in vision-based systems—such as autonomous driving—to adversarial attacks, particularly under resource-constrained deployment scenarios where existing defenses often incur prohibitive computational overhead. The authors propose a lightweight adversarial example detection mechanism grounded in multi-task consistency, leveraging discrepancies between outputs of complementary tasks like instance segmentation and object detection to construct a consistency-based scoring metric. An efficient model-pair selection strategy is further introduced to enhance practicality. Integrated within a multi-task learning framework, the method achieves a 99.9% ROC-AUC detection performance on the BDD100k dataset, demonstrating substantial improvements in both efficiency and robustness across diverse vision models and under PGD attacks.

Abstract

Deep Neural Networks (DNNs) have found successful deployment in numerous vision perception systems. However, their susceptibility to adversarial attacks has prompted concerns regarding their practical applications, specifically in the context of autonomous driving. Existing defenses often suffer from cost inefficiency, rendering their deployment impractical for resource-constrained applications. In this work, we propose an efficient and effective adversarial attack detection scheme leveraging the multi-task perception within a complex vision system. Adversarial perturbations are detected by the inconsistencies between the inference outputs of multiple vision tasks, e.g., object detection and instance segmentation. To this end, we developed a consistency score metric to measure the inconsistency between vision tasks. Next, we designed an approach to select the best model pairs for detecting inconsistencies effectively. Finally, we evaluated our defense against PGD attacks across multiple vision models on the BDD100k validation dataset. The experimental results demonstrated that our defense achieved a ROC-AUC performance of 99.9% detection within the considered attacker model.

Problem

• adversarial attacks
• deep neural networks
• multi-task perception
• autonomous driving
• defense efficiency

Innovation

• multi-task consistency
• adversarial attack detection
• vision perception systems
• consistency score
• model pair selection