Recent approaches employ deep learning-based solutions for the recovery of a sharp image from its blurry observation. This paper introduces adversarial attacks against deep learning-based image deblurring methods and evaluates the robustness of these neural networks to untargeted and targeted attacks. We demonstrate that imperceptible distortion can significantly degrade the performance of state-of-the-art deblurring networks, even producing drastically different content in the output, indicating the strong need to include adversarially robust training not only in classification but also for image recovery.
@article{arxiv.2210.02502,
title = {On Adversarial Robustness of Deep Image Deblurring},
author = {Kanchana Vaishnavi Gandikota and Paramanand Chandramouli and Michael Moeller},
journal= {arXiv preprint arXiv:2210.02502},
year = {2022}
}