Research map / Adversarial robustness and security / Attacks on text and graphs
Towards Defending against Adversarial Examples via Attack-Invariant Features
Dawei Zhou, Tongliang Liu, Bo Han, Nannan Wang, Chunlei Peng, Xinbo Gao
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 |
|---|---|---|---|
| Use ARN instead of HGD_PP | LeNet · MNIST, JSMA_T | classification error rate (%): 50.62 -> 16.75 | Improved |
| Use ARN_PP instead of HGD_PP | ResNet-110 · CIFAR-10, PGD_N | classification error rate (%): 39.44 -> 38.66 | Improved |
| Use ARN instead of APE-G | LeNet · MNIST, BPDA+PGD_N, Ite-40 | classification error rate (%): 72.01 -> 24.65 | Improved |
| Use ARN instead of APE-G | ResNet-110 · CIFAR-10, BPDA+PGD_N, Ite-40 | classification error rate (%): 89.06 -> 60.47 | Improved |
| Add pixel-space adversarial loss L_adv (ablation) | ARN · CIFAR-10, attacks including STA_N | classification accuracy rate (%) | Improved |
| Add normalization loss L_nor (ablation) | ARN · CIFAR-10, STA_N | classification accuracy rate (%) | Improved |
| Add attack-invariant loss L_att (ablation) | ARN · CIFAR-10, unseen AA_N, PGD_Nε', and STA_N | classification accuracy rate (%) | Improved |
Closest papers by meaning
- Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks (CVPR 2019 · similarity 0.91)
- Fundamental Tradeoffs between Invariance and Sensitivity to Adversarial Perturbations (ICML 2020 · similarity 0.89)
- Towards Deep Learning Models Resistant to Adversarial Attacks (ICLR 2018 · similarity 0.88)
- Breaking Barriers in Physical-World Adversarial Examples: Improving Robustness and Transferability via Robust Feature (AAAI 2025 · similarity 0.87)
- Wide Two-Layer Networks can Learn from Adversarial Perturbations (NeurIPS 2024 · similarity 0.87)
- Understanding the Robustness of Randomized Feature Defense Against Query-Based Adversarial Attacks (ICLR 2024 · similarity 0.87)
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.
