Towards Understanding Adversarial Robustness of Optical Flow Networks
This addresses security concerns for automotive systems by improving adversarial robustness in optical flow estimation, though it is incremental as it builds on known attack methods.
The paper analyzes the vulnerability of optical flow networks to physical patch-based adversarial attacks, attributing it to the aperture problem and architectural flaws, and proposes rectifications to enhance robustness.
Recent work demonstrated the lack of robustness of optical flow networks to physical patch-based adversarial attacks. The possibility to physically attack a basic component of automotive systems is a reason for serious concerns. In this paper, we analyze the cause of the problem and show that the lack of robustness is rooted in the classical aperture problem of optical flow estimation in combination with bad choices in the details of the network architecture. We show how these mistakes can be rectified in order to make optical flow networks robust to physical patch-based attacks. Additionally, we take a look at global white-box attacks in the scope of optical flow. We find that targeted white-box attacks can be crafted to bias flow estimation models towards any desired output, but this requires access to the input images and model weights. However, in the case of universal attacks, we find that optical flow networks are robust. Code is available at https://github.com/lmb-freiburg/understanding_flow_robustness.