Research map / Label-efficient and robust learning / Domain adaptation and generalization
Open Domain Generalization with Domain-Augmented Meta-Learning
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 |
|---|---|---|---|
| Use DAML instead of Epi-FCR | Open-domain generalization · PACS (Avg) | Acc: 60.64\pm{0.22} -> 65.49\pm{0.36} | Helped |
| Use DAML instead of Epi-FCR | Open-domain generalization · PACS (Avg) | H-score: 48.47\pm{0.29} -> 51.88\pm{0.42} | Helped |
| Use DAML instead of CuMix | Open-domain generalization · Office-Home (Avg) | Acc: 51.67\pm{0.12} -> 56.45\pm{0.21} | Helped |
| Use DAML instead of CuMix | Open-domain generalization · Office-Home (Avg) | H-score: 49.40\pm{0.27} -> 53.34\pm{0.45} | Helped |
| Use DAML instead of CuMix | Open-domain generalization · Multi-Datasets (Avg) | Acc: 42.18\pm{0.45} -> 46.61\pm{0.59} | Helped |
| Use DAML instead of CuMix | Open-domain generalization · Multi-Datasets (Avg) | H-score: 46.91\pm{0.40} -> 51.71\pm{0.52} | Helped |
| Use DAML instead of CuMix | Closed-set generalization · PACS (Avg) | accuracy: 81.6 -> 82.7 | Helped |
| Add Dir-mixup in meta-training (ablation) | DAML · Office-Home open-domain (Avg) | Acc: 53.6 -> 54.4 | Helped |
| Add Dir-mixup in the meta-objective (ablation) | DAML · Office-Home open-domain (Avg) | Acc: 53.6 -> 55.0 | Helped |
| Use Dir-mixup in both meta-training and the meta-objective (ablation) | DAML · Office-Home open-domain (Avg) | Acc: 54.4 -> 55.9; 55.0 -> 55.9 | Helped |
| Add distilled soft-labeling (ablation) | DAML with both Dir-mixup terms · Office-Home open-domain (Avg) | Acc: 55.9 -> 56.5 | Helped |
| Add meta-learning rather than supervised learning on augmented domains (ablation) | DAML · Office-Home open-domain (Avg) | Acc: 55.3 -> 56.5 | Helped |
Closest papers by meaning
- Domain Generalization via Model-Agnostic Learning of Semantic Features (NeurIPS 2019 · similarity 0.91)
- MetaReg: Towards Domain Generalization using Meta-Regularization (NeurIPS 2018 · similarity 0.91)
- Bi-Level Meta-Learning for Few-Shot Domain Generalization (CVPR 2023 · similarity 0.91)
- Multimodal Unsupervised Domain Generalization by Retrieving Across the Modality Gap (ICLR 2025 · similarity 0.88)
- PracticalDG: Perturbation Distillation on Vision-Language Models for Hybrid Domain Generalization (CVPR 2024 · similarity 0.87)
- Towards Generalizing to Unseen Domains with Few Labels (CVPR 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.