Universal Robustness via Median Randomized Smoothing for Real-World Super-Resolution

TL;DR

This paper demonstrates that median randomized smoothing (MRS) provides a more universal and adaptable robustness method for real-world super-resolution models, outperforming adversarial training across various corruptions and attack types.

Contribution

The paper introduces median randomized smoothing as a universal robustness technique for super-resolution, showing its superior generalization over adversarial methods in diverse real-world scenarios.

Findings

MRS outperforms adversarial training in robustness across attack types.

MRS effectively handles standard and real-world image corruptions.

Universal robustness via MRS improves real-world super-resolution performance.

Abstract

Most of the recent literature on image Super-Resolution (SR) can be classified into two main approaches. The first one involves learning a corruption model tailored to a specific dataset, aiming to mimic the noise and corruption in low-resolution images, such as sensor noise. However, this approach is data-specific, tends to lack adaptability, and its accuracy diminishes when faced with unseen types of image corruptions. A second and more recent approach, referred to as Robust Super-Resolution (RSR), proposes to improve real-world SR by harnessing the generalization capabilities of a model by making it robust to adversarial attacks. To delve further into this second approach, our paper explores the universality of various methods for enhancing the robustness of deep learning SR models. In other words, we inquire: "Which robustness method exhibits the highest degree of adaptability when dealing with a wide range of adversarial attacks ?". Our extensive experimentation on both synthetic and real-world images empirically demonstrates that median randomized smoothing (MRS) is more general in terms of robustness compared to adversarial learning techniques, which tend to focus on specific types of attacks. Furthermore, as expected, we also illustrate that the proposed universal robust method enables the SR model to handle standard corruptions more effectively, such as blur and Gaussian noise, and notably, corruptions naturally present in real-world images. These results support the significance of shifting the paradigm in the development of real-world SR methods towards RSR, especially via MRS.