Research map / Adversarial robustness and security / Attacks on text and graphs
Attack to Explain Deep Representation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use distribution-level model-explaining perturbations instead of the comparison attack | Robust ResNet image generation · ImageNet image generation | visual image quality | Helped |
| Use the proposed attack instead of the comparison method | Robust ResNet inpainting · Image inpainting | visual restoration quality | Helped |
| Use model-explaining perturbations and refinement instead of the comparison method | Robust-classifier image manipulation · Interactive image manipulation | visual realism | Helped |
| Accumulate targeted perturbations across an image distribution rather than compute an image-specific attack | Standard deep visual classifiers · ImageNet target-label perturbations | presence of salient target-class features | Helped |
| Change the seed image (ablation) | Proposed attack · ImageNet target-label perturbations | diversity of salient patterns | Mixed |
| Constrain corresponding image regions with uniform patches (ablation) | Proposed attack · Centipede target-label perturbations | focus near the specified region | Helped |
| Apply the method to different classifier architectures (ablation) | Proposed attack · ImageNet target-label perturbations | presence of salient target-class features | No clear effect |
Closest papers by meaning
- Generative Adversarial Perturbations (CVPR 2018 · similarity 0.89)
- PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examples (ICLR 2018 · similarity 0.88)
- Modeling Biological Immunity to Adversarial Examples (CVPR 2020 · similarity 0.88)
- Theoretical Understanding of Learning from Adversarial Perturbations (ICLR 2024 · similarity 0.87)
- Wide Two-Layer Networks can Learn from Adversarial Perturbations (NeurIPS 2024 · similarity 0.87)
- Pixel2Feature Attack (P2FA): Rethinking the Perturbed Space to Enhance Adversarial Transferability (ICML 2025 · similarity 0.86)
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.