Research map / Label-efficient and robust learning / Domain adaptation and generalization
Adversarially Adaptive Normalization for Single Domain Generalization
Check 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 |
|---|---|---|---|
| Add ASR-Norm to RSDA | RSDA · Digits (MNIST source; target domains) | average accuracy: 68.5 -> 70.1 | Helped |
| Add ASR-Norm to ADA | ADA · CIFAR-10-C | average accuracy: 75.7 -> 86.0 | Helped |
| Add ASR-Norm to RSC | RSC · PACS (one source domain) | average accuracy: 61.8 -> 68.1 | Helped |
| Add ASR-Norm to RSC | RSC · PACS (three source domains; no domain labels) | average accuracy: 85.2 -> 86.3 | Helped |
| Add adaptive rescaling to adaptive standardization (ablation) | ADA+AS · CIFAR-10-C | average accuracy: 85.4 -> 86.0 | Helped |
| Add adaptive standardization to adaptive rescaling (ablation) | ADA+AR · CIFAR-10-C | average accuracy: 84.2 -> 86.0 | Helped |
| Use adaptive standardization instead of switchable normalization (ablation) | ADA with normalization · CIFAR-10-C | average accuracy: 84.4 -> 85.4 | Helped |
| Apply ASR-Norm to more normalization layers (ablation) | ASR-Norm on ResNet-18 · PACS (one source domain) | average accuracy: 62.3 -> 64.4 | Helped |
| Increase C_stan from C/16 to C/8 for all ASR-Norm layers (ablation) | ASR-Norm on ResNet-18 · PACS (one source domain) | average accuracy: 66.5 -> 67.5 | Helped |
| Increase C_stan from C/8 to C/2 for all ASR-Norm layers (ablation) | ASR-Norm on ResNet-18 · PACS (one source domain) | average accuracy: 67.5 -> 68.1 | Helped |
| Use ASR-Norm rather than BN (ablation) | ADA with normalization · CIFAR-10 test | accuracy: 95.4 -> 94.6 | Hurt |
| Use ASR-Norm rather than BN (ablation) | ADA with normalization · CIFAR-10 adversarially corrupted images | accuracy: 32.0 -> 52.2 | Helped |
Closest papers by meaning
- Learning to Learn Single Domain Generalization (CVPR 2020 · similarity 0.91)
- Domain Adaptation with Adversarial Training on Penultimate Activations (AAAI 2023 · similarity 0.90)
- Sharpness-Aware Model-Agnostic Long-Tailed Domain Generalization (AAAI 2024 · similarity 0.89)
- When Neural Networks Fail to Generalize? A Model Sensitivity Perspective (AAAI 2023 · similarity 0.89)
- Reducing Divergence in Batch Normalization for Domain Adaptation (AAAI 2025 · similarity 0.88)
- Adversarial Domain Prompt Tuning and Generation for Single Domain Generalization (CVPR 2025 · 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.