Research map / Label-efficient and robust learning / Domain adaptation and generalization
Combating Label Distribution Shift for Active Domain Adaptation
Sehyun Hwang, Sohyun Lee, Sungyeon Kim, Jungseul Ok, Suha Kwak
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 |
|---|---|---|---|
| add MMD prototype sampling instead of random sampling (ablation) | DANN · OfficeHome | accuracy (%): 73.3 -> 77.1 | Improved |
| add label distribution matching (ablation) | LAMDA (prototype, linear) · OfficeHome-RSUT | accuracy (%): 67.3 -> 74.1 | Improved |
| add label distribution matching (ablation) | LAMDA (prototype, linear) · OfficeHome | accuracy (%): 77.1 -> 79.8 | Improved |
| add cosine classifier (ablation) | LAMDA · OfficeHome | accuracy (%): 79.8 -> 81.9 | Improved |
| add cosine classifier (ablation) | LAMDA · OfficeHome-RSUT | accuracy (%): 74.1 -> 75.8 | Improved |
| add cosine classifier (ablation) | LAMDA · VisDA-2017 | accuracy (%): 92.3 -> 91.8 | Hurt |
| add cosine classifier (ablation) | LAMDA · DomainNet | accuracy (%): 57.1 -> 57.0 | No clear effect |
| add uncertainty-based labeling of hard prototypes only (ablation) | LAMDA sampling · OfficeHome | accuracy (%): 77.9 -> 80.3 | Improved |
| use pseudo-labeled prototypes in label distribution estimation (ablation) | LAMDA sampling · OfficeHome | accuracy (%): 80.3 -> 82.0 | Improved |
| add MMD density-aware sampling instead of random (ablation) | DANN · OfficeHome | accuracy (%): 73.3 -> 77.9 | Improved |
| use pseudo labels (Delta below 1) instead of Delta = 1 (ablation) | LAMDA · OfficeHome and OfficeHome-RSUT | accuracy (%): drops 4% and 1.1% | Improved |
| use LAMDA sampling instead of CLUE under MCC | MCC · OfficeHome | accuracy (%): 83.2 -> 84.2 | Improved |
Closest papers by meaning
- Divide and Adapt: Active Domain Adaptation via Customized Learning (CVPR 2023 · similarity 0.90)
- Category-Aware Active Domain Adaptation (ICML 2024 · similarity 0.90)
- Revisiting the Domain Shift and Sample Uncertainty in Multi-source Active Domain Transfer (CVPR 2024 · similarity 0.90)
- Enhancing Semi-supervised Domain Adaptation via Effective Target Labeling (AAAI 2024 · similarity 0.90)
- Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.89)
- Diffusion-Driven Progressive Target Manipulation for Source-Free Domain Adaptation (NeurIPS 2025 · similarity 0.89)
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.
