Research map / Image generation / Super-resolution
Evaluating Robustness of Deep Image Super-Resolution Against Adversarial Attacks
Jun-Ho Choi, Huan Zhang, Jun-Hyuk Kim, Cho-Jui Hsieh, Jong-Seok Lee
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the resizing pre-processing defense (shrink one pixel, resize back) | EDSR under basic attack · Set5 (alpha=8/255) | PSNR of SR output (dB): 16.14 -> 25.01 | Improved |
| add geometric self-ensemble as defense | EDSR under basic attack · Set5 (alpha=8/255) | PSNR of SR output (dB): 16.14 -> 23.47 | Improved |
| train with GAN/perceptual objective (ESRGAN vs RRDB, same architecture) | super-resolution models · BSD100 basic attack | PSNR of SR output under attack | Hurt |
| use a smaller model (EDSR-baseline instead of EDSR) | EDSR · BSD100 / Set5 basic attack | PSNR of SR output under attack | Improved |
| replace deep SR model with bicubic interpolation | SR models · Set5, Set14, BSD100 basic attack | robustness to attack | Improved |
| switch from image-specific to universal attack | SR attack · BSD100 | PSNR of SR output under attack | Hurt |
Closest papers by meaning
- Universal Robustness via Median Randomized Smoothing for Real-World Super-Resolution (CVPR 2024 · similarity 0.91)
- AND: Adversarial Neural Degradation for Learning Blind Image Super-Resolution (NeurIPS 2023 · similarity 0.87)
- Resolution Attack: Exploiting Image Compression to Deceive Deep Neural Networks (ICLR 2025 · similarity 0.87)
- Fast Bayesian Uncertainty Estimation and Reduction of Batch Normalized Single Image Super-Resolution Network (CVPR 2021 · similarity 0.86)
- SRFeat: Single Image Super-Resolution with Feature Discrimination (ECCV 2018 · similarity 0.86)
- Uncertainty-Aware GAN for Single Image Super Resolution (AAAI 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
