Adversarial Training of Self-supervised Monocular Depth Estimation against Physical-World Attacks
This work addresses security threats in MDE systems, particularly for autonomous driving, by enhancing robustness against adversarial attacks in a self-supervised setting, representing an incremental improvement over existing methods.
The paper tackles the problem of adversarial robustness in self-supervised monocular depth estimation (MDE) for applications like autonomous driving, proposing a novel adversarial training method that improves robustness against physical-world attacks with nearly no degradation in benign performance.
Monocular Depth Estimation (MDE) is a critical component in applications such as autonomous driving. There are various attacks against MDE networks. These attacks, especially the physical ones, pose a great threat to the security of such systems. Traditional adversarial training method requires ground-truth labels hence cannot be directly applied to self-supervised MDE that does not have ground-truth depth. Some self-supervised model hardening techniques (e.g., contrastive learning) ignore the domain knowledge of MDE and can hardly achieve optimal performance. In this work, we propose a novel adversarial training method for self-supervised MDE models based on view synthesis without using ground-truth depth. We improve adversarial robustness against physical-world attacks using L0-norm-bounded perturbation in training. We compare our method with supervised learning based and contrastive learning based methods that are tailored for MDE. Results on two representative MDE networks show that we achieve better robustness against various adversarial attacks with nearly no benign performance degradation.