Research map / Label-efficient and robust learning / Domain adaptation and generalization
A Fourier-Based Framework for 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 |
|---|---|---|---|
| Use FACT rather than L2A-OT | Digits-DG · Digits-DG leave-one-domain-out | average accuracy: 78.1 -> 81.5 | Helped |
| Use FACT rather than RSC | PACS with ResNet50 · PACS leave-one-domain-out | average accuracy: 87.83 -> 88.15 | Helped |
| Use FACT rather than L2A-OT | OfficeHome · OfficeHome leave-one-domain-out | average accuracy: 65.60 -> 66.56 | Helped |
| Add amplitude mix to the baseline (ablation) | PACS with ResNet18 · PACS leave-one-domain-out | average accuracy: 79.94 -> 83.44 | Helped |
| Add dual consistency losses to amplitude mix (ablation) | PACS with ResNet18 · PACS leave-one-domain-out | average accuracy: 83.44 -> 83.71 | Helped |
| Add a momentum-updated teacher to dual consistency (ablation) | PACS with ResNet18 · PACS leave-one-domain-out | average accuracy: 83.71 -> 84.51 | Helped |
| Add amplitude mix to co-teacher regularization (ablation) | PACS with ResNet18 · PACS leave-one-domain-out | average accuracy: 82.71 -> 84.51 | Helped |
| Use both directions of co-teacher consistency instead of a2o only (ablation) | PACS with ResNet18 · PACS leave-one-domain-out | average accuracy: 83.67 -> 84.51 | Helped |
| Use both directions of co-teacher consistency instead of o2a only (ablation) | PACS with ResNet18 · PACS leave-one-domain-out | average accuracy: 83.84 -> 84.51 | Helped |
| Replace partial amplitude swap with amplitude mix (ablation) | DeepAll · PACS leave-one-domain-out | average accuracy: 82.34 -> 83.44 | Helped |
| Replace partial amplitude swap with amplitude mix (ablation) | FACT · PACS leave-one-domain-out | average accuracy: 82.67 -> 84.51 | Helped |
| Use amplitude mix rather than amplitude elimination (ablation) | FACT · PACS leave-one-domain-out | average accuracy: 82.74 -> 84.51 | Helped |
Closest papers by meaning
- Domain Generalisation via Domain Adaptation: An Adversarial Fourier Amplitude Approach (ICLR 2023 · similarity 0.91)
- Domain Generalization with Vital Phase Augmentation (AAAI 2024 · similarity 0.86)
- Seeking Consistent Flat Minima for Better Domain Generalization via Refining Loss Landscapes (CVPR 2025 · similarity 0.86)
- Towards Combating Frequency Simplicity-biased Learning for Domain Generalization (NeurIPS 2024 · similarity 0.85)
- Gradient-Guided Annealing for Domain Generalization (CVPR 2025 · similarity 0.85)
- Disentangled Prompt Representation for Domain Generalization (CVPR 2024 · 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.