Research map / Label-efficient and robust learning / Domain adaptation and generalization
Learning to Learn 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 Wasserstein relaxation to adversarial domain augmentation (ablation) | M-ADA · Digits | mean accuracy: 55.43 -> 59.49 | Helped |
| Add embedding-space consistency loss (ablation) | M-ADA · Digits | mean accuracy: 58.70 -> 59.49 | Helped |
| Add meta-learning (ablation) | M-ADA · Digits | mean accuracy: 58.55 -> 59.49 | Helped |
| Add Wasserstein relaxation to adversarial domain augmentation (ablation) | M-ADA · CIFAR-10-C severity 5 | average accuracy: 61.92 -> 65.59 | Helped |
| Add meta-learning (ablation) | M-ADA · CIFAR-10-C severity 5 | average accuracy: 64.22 -> 65.59 | Helped |
| Use WAE rather than Vanilla AE for the relaxation term (ablation) | M-ADA · Digits | accuracy: 58.67% -> 59.49% | Helped |
| Use WAE rather than input-distance relaxation (ablation) | M-ADA · CIFAR-10-C | accuracy: 62.03% -> 65.59% | Helped |
| Set the number of augmented domains to K=3 rather than increasing K further (ablation) | M-ADA · Digits and CIFAR-10-C | accuracy | Mixed |
| Increase the embedding-distance constraint weight alpha beyond 1.0 (ablation) | M-ADA · Digits and CIFAR-10-C | accuracy | Hurt |
| Increase the domain-deviation weight beta beyond 2.0 × 10^3 (ablation) | M-ADA · Digits and CIFAR-10-C | accuracy | Hurt |
| Use M-ADA rather than GUD | Single-domain generalization · Digits | mean accuracy: 54.62 -> 59.49 | Helped |
| Use M-ADA rather than GUD | Single-domain generalization · CIFAR-10-C severity 5 | average accuracy: 58.26 -> 65.59 | Helped |
Closest papers by meaning
- Domain Generalization With Adversarial Feature Learning (CVPR 2018 · similarity 0.92)
- Adversarially Adaptive Normalization for Single Domain Generalization (CVPR 2021 · similarity 0.91)
- Generalizing to Unseen Domains via Adversarial Data Augmentation (NeurIPS 2018 · similarity 0.91)
- Sharpness-Aware Model-Agnostic Long-Tailed Domain Generalization (AAAI 2024 · similarity 0.90)
- Adversarial Domain Prompt Tuning and Generation for Single Domain Generalization (CVPR 2025 · similarity 0.90)
- AdvST: Revisiting Data Augmentations for Single Domain Generalization (AAAI 2024 · 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.