Research map / Adversarial robustness and security / Certified and empirical robustness
Optimal Transport Model Distributional Robustness
Van-Anh Nguyen, Trung Le, Anh Tuan Bui, Thanh-Toan Do, Dinh Phung
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 |
|---|---|---|---|
| replace SAM with OT-MDR | WideResNet28x10 · CIFAR-100 | accuracy: 82.69 -> 84.14 | Improved |
| replace SAM with OT-MDR | Densenet121 · CIFAR-10 | accuracy: 91.16 -> 91.44 | Improved |
| replace bSAM with OT-MDR | ResNet18 · CIFAR-100 | accuracy: 80.22 -> 81.23 | Improved |
| replace bSAM with OT-MDR | ResNet18 · CIFAR-100 | AUROC: 0.892 -> 0.991 | Improved |
| replace SAM with OT-MDR in ensemble | five Resnet10 ensemble · CIFAR-10 | accuracy: 95.4 -> 95.4 | No clear effect |
| replace SAM with OT-MDR in ensemble | five Resnet10 ensemble · CIFAR-10 | ECE: 0.050 -> 0.021 | Improved |
| replace SAM with OT-MDR in ensemble | five Resnet10 ensemble · CIFAR-100 | ECE: 0.136 -> 0.043 | Improved |
| replace SAM with OT-MDR in heterogeneous ensemble | ResNet18+MobileNet+EfficientNet · CIFAR-100 | accuracy: 76.4 -> 78.3 | Improved |
| add OT-MDR to SGVB | Bayesian NN Resnet10 · CIFAR-10 | NLL: 0.78 -> 0.81 | Hurt |
| add OT-MDR to SGVB | Bayesian NN Resnet18 · CIFAR-100 | accuracy: 60.91 -> 63.17 | Improved |
| add OT-MDR to SGVB | Bayesian NN Resnet10 · CIFAR-10 | ECE: 0.23 -> 0.26 | Hurt |
| increase number of particles from 1 to 2 (ablation) | OT-MDR WideResNet28x10 · CIFAR | accuracy | Improved |
Closest papers by meaning
- Bootstrap Your Uncertainty: Adaptive Robust Classification Driven by Optimal-Transport (NeurIPS 2025 · similarity 0.87)
- Distributional Robustness with IPMs and links to Regularization and GANs (NeurIPS 2020 · similarity 0.87)
- Outlier-Robust Distributionally Robust Optimization via Unbalanced Optimal Transport (NeurIPS 2024 · similarity 0.87)
- Universal generalization guarantees for Wasserstein distributionally robust models (ICLR 2025 · similarity 0.86)
- Diffusion Models are Certifiably Robust Classifiers (NeurIPS 2024 · similarity 0.85)
- Structure-informed Risk Minimization for Robust Ensemble Learning (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.
