Research map / Adversarial robustness and security / Certified and empirical robustness
Improving Adversarial Robustness via Mutual Information Estimation
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 NAMID's natural/adversarial MI objective with standard adversarial training | ResNet-18 · CIFAR-10, white-box PGD-40 | adversarial accuracy: 5.69% | Helped |
| Use NAMID's natural/adversarial MI objective | ResNet-18 · CIFAR-10, white-box attacks | adversarial accuracy | Helped |
| Use NAMID's natural/adversarial MI objective | ResNet-18 · Tiny-ImageNet, white-box attacks | adversarial accuracy | Helped |
| Use NAMID's natural/adversarial MI objective | ResNet-18 · CIFAR-10 and Tiny-ImageNet, black-box attacks | adversarial accuracy | Helped |
| Replace standard-MI maximization with the proposed natural/adversarial MI estimator training mechanism | NAMID MI estimators · CIFAR-10 adversarial defense | adversarial accuracy | Helped |
| Remove the adversarial MI term (ablation) | NAMID · L∞ PGD and FWA attacks | adversarial accuracy | Hurt |
| Remove the natural MI term (ablation) | NAMID · L∞ PGD and FWA attacks | adversarial accuracy | Hurt |
| Set λ to 0 (ablation) | NAMID · L∞ PGD and FWA attacks | adversarial accuracy | Hurt |
Closest papers by meaning
- Towards Deep Learning Models Resistant to Adversarial Attacks (ICLR 2018 · similarity 0.89)
- Phase-aware Adversarial Defense for Improving Adversarial Robustness (ICML 2023 · similarity 0.89)
- Increasing Confidence in Adversarial Robustness Evaluations (NeurIPS 2022 · similarity 0.89)
- Adversarial Training Can Provably Improve Robustness: Theoretical Analysis of Feature Learning Process Under Structured Data (ICLR 2025 · similarity 0.89)
- Improving Adversarial Robustness by Putting More Regularizations on Less Robust Samples (ICML 2023 · similarity 0.89)
- Error Correcting Output Codes Improve Probability Estimation and Adversarial Robustness of Deep Neural Networks (NeurIPS 2019 · similarity 0.88)
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.