Research map / Adversarial robustness and security / Black-box and transfer attacks
EAD: Elastic-Net Attacks to Deep Neural Networks via Adversarial Examples
Read the paperCheck 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 elastic-net EAD with the EN decision rule rather than C&W | targeted attacks on MNIST · MNIST average case | L1 distortion: 22.46 -> 17.4 | Helped |
| Use elastic-net EAD with the EN decision rule rather than C&W | targeted attacks on CIFAR10 · CIFAR10 average case | L1 distortion: 13.62 -> 8.18 | Helped |
| Use elastic-net EAD with the EN decision rule rather than C&W | targeted attacks on ImageNet · ImageNet average case | L1 distortion: 232.2 -> 69.47 | Helped |
| Use the L1 decision rule instead of the EN decision rule (ablation) | EAD on MNIST · MNIST average case | L1 distortion: 17.4 -> 14.11 | Helped |
| Use the L1 decision rule instead of the EN decision rule (ablation) | EAD on CIFAR10 · CIFAR10 average case | L1 distortion: 8.18 -> 6.066 | Helped |
| Use the L1 decision rule instead of the EN decision rule (ablation) | EAD on ImageNet · ImageNet average case | L1 distortion: 69.47 -> 40.9 | Helped |
| Use the L1 decision rule instead of the EN decision rule (ablation) | EAD on MNIST · MNIST average case | L2 distortion: 2.001 -> 2.211 | Hurt |
| Remove the L1 penalty by setting beta to 0 (ablation) | EAD on MNIST · MNIST average case | L1 distortion: 17.4 -> 22.63 | Hurt |
| Increase beta from 0 to 10^-2 (ablation) | EAD on MNIST · MNIST average case | L1 distortion: 22.63 -> 13.56 | Helped |
| Use COV optimization instead of EAD's shrinkage-thresholding optimization (ablation) | elastic-net attack on MNIST · MNIST average case | L1 distortion: 17.4 (EAD, beta 10^-3) vs 22.42 (COV, beta 10^-3) | Hurt |
Closest papers by meaning
- On Adversarial Training without Perturbing all Examples (ICLR 2024 · similarity 0.87)
- Adversarial Example Games (NeurIPS 2020 · similarity 0.87)
- NATTACK: Learning the Distributions of Adversarial Examples for an Improved Black-Box Attack on Deep Neural Networks (ICML 2019 · similarity 0.87)
- Beyond ImageNet Attack: Towards Crafting Adversarial Examples for Black-box Domains (ICLR 2022 · similarity 0.87)
- Enhancing Adversarial Transferability with Adversarial Weight Tuning (AAAI 2025 · similarity 0.86)
- DiffAdvMAP: Flexible Diffusion-Based Framework for Generating Natural Unrestricted Adversarial Examples (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.