Research map / Adversarial robustness and security / Certified and empirical robustness
CERTIFIED VS. EMPIRICAL ADVERSARIAL ROBUSTNESS VIA HYBRID CONVOLUTIONS WITH ATTENTION STOCHASTICITY
Joy Dhar, Song Xia, Manish Kumar Pandey, Maryam Haghighat, Azadeh Alavi, Ferdous Sohel, Wenyu Zhang, Nayyar Zaidi
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 |
|---|---|---|---|
| increase smoothing noise sigma from 0.25 to 0.50 in HyCAS (ablation) | HyCAS · CIFAR-10, radius 2.0 | certified accuracy (%): 8.52 -> 12.5 | Improved |
| increase smoothing noise sigma from 0.25 to 0.50 in HyCAS (ablation) | HyCAS · ImageNet, radius 2.0 | certified accuracy (%): 5.42 -> 24.8 | Improved |
| increase smoothing noise sigma from 0.25 to 0.50 in HyCAS (ablation) | HyCAS · CIFAR-10, radius 0.75 | certified accuracy (%): 44.3 -> 44.3 | No clear effect |
| use HyCAS instead of ARS (hybrid Lipschitz convolutions plus attention noise) | randomized smoothing certification · CIFAR-10, radius 0.75, sigma 0.25 | certified accuracy (%): 39.1 -> 44.3 | Improved |
| use HyCAS instead of CTRW | empirical l-inf defense · NIH-CXR, APGD-20 eps 8/255 | robust accuracy (%): 85.1 -> 88.6 | Improved |
| use HyCAS instead of CTRW | empirical l-inf defense · HAM10000, APGD-20 eps 8/255 | robust accuracy (%): 64.7 -> 67.8 | Improved |
| use HyCAS instead of CTRW | empirical l-inf defense · EyePACS, APGD-20 eps 8/255 | robust accuracy (%): 70.1 -> 72.6 | Improved |
| increase APGD attack steps from 10 to 100 | HyCAS vs prior defenses · CIFAR-100, APGD steps | robust accuracy: HyCAS stays 7-12% above closest competitor at 100 steps | Mixed |
| increase certified radius (via larger smoothing noise) | HyCAS · CIFAR-10 and ImageNet | empirical robustness | Hurt |
Closest papers by meaning
- Does a Hybrid Space-Aware Randomized Defense Improve Empirical and Certified Adversarial Robustness? (ICML 2026 · similarity 0.94)
- A Combination of Noise and Bilateral Filters Achieve Supralinear and Scalable Adversarial Robustness in CNNs (CVPR 2026 · similarity 0.88)
- Adversarial Robustness via Deformable Convolution with Stochasticity (ICML 2025 · similarity 0.87)
- A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking (IJCV 2024 · similarity 0.87)
- (Certified!!) Adversarial Robustness for Free! (ICLR 2023 · similarity 0.87)
- ROME is Forged in Adversity: Robust Distilled Datasets via Information Bottleneck (ICML 2025 · 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.
