Fine-Tuning Adversarially-Robust Transformers for Single-Image Dehazing
This addresses the reliability problem for remote sensing applications by enhancing adversarial robustness in single-image dehazing, though it is incremental as it builds on existing transformer methods.
The paper tackled the susceptibility of state-of-the-art image-to-image dehazing transformers to adversarial noise, showing that even a 1-pixel change can reduce PSNR by up to 2.8 dB, and proposed lightweight fine-tuning strategies that maintain clean performance while significantly increasing robustness against such attacks.
Single-image dehazing is an important topic in remote sensing applications, enhancing the quality of acquired images and increasing object detection precision. However, the reliability of such structures has not been sufficiently analyzed, which poses them to the risk of imperceptible perturbations that can significantly hinder their performance. In this work, we show that state-of-the-art image-to-image dehazing transformers are susceptible to adversarial noise, with even 1 pixel change being able to decrease the PSNR by as much as 2.8 dB. Next, we propose two lightweight fine-tuning strategies aimed at increasing the robustness of pre-trained transformers. Our methods results in comparable clean performance, while significantly increasing the protection against adversarial data. We further present their applicability in two remote sensing scenarios, showcasing their robust behavior for out-of-distribution data. The source code for adversarial fine-tuning and attack algorithms can be found at github.com/Vladimirescu/RobustDehazing.