# Enhancing Robustness in UDC Image Restoration Through Adversarial Purification and Fine-Tuning

## TL;DR

This paper introduces a new framework to protect UDC image restoration models from adversarial attacks, significantly improving their robustness and performance.

## Contribution

A novel two-stage defense framework combining adversarial purification and fine-tuning for UDC image restoration is proposed and validated.

## Key findings

- Seven UDC models showed severe performance degradation under adversarial attacks, with DISCNet's PSNR dropping from 35.24 to 15.16 under C&W attack.
- The proposed framework improved DISCNet's PSNR to 32.17 under PGD attack and UFormer's PSNR to 19.71 under LPIPS-guided attacks.
- Extensive experiments confirmed the framework's effectiveness in enhancing resilience against various adversarial attacks.

## Abstract

This study presents a novel defense framework to fortify Under-Display Camera (UDC) image restoration models against adversarial attacks, a previously underexplored vulnerability in this domain. Our research initially conducts an in-depth robustness evaluation of deep-learning-based UDC image restoration models by employing several white-box and black-box attacking methods. Following the assessment, we propose a two-stage approach integrating diffusion-based adversarial purification and efficient fine-tuning, uniquely designed to eliminate perturbations while retaining restoration fidelity. For the first time, we systematically evaluate seven state-of-the-art UDC models (such as DISCNet, UFormer, etc.) under diverse attacks (PGD, C&W, etc.), revealing severe performance degradation (DISCNet’s PSNR drops from 35.24 to 15.16 under C&W attack). Our framework demonstrates significant improvements: after purification and fine-tuning, DISCNet’s PSNR rebounds to 32.17 under PGD attack (vs. 30.17 without defense), while UFormer achieves a 19.71 PSNR under LPIPS-guided attacks (vs. 17.38 baseline). The effectiveness of our proposed approach is validated through extensive experiments, showing marked improvements in resilience against various adversarial attacks.