Research map / Label-efficient and robust learning / Domain adaptation and generalization
Confidence Score for Source-Free Unsupervised Domain Adaptation
Read the paperCheck 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 CoWA-JMDS instead of NRC | SFUDA · Office-31 closed-set, Avg. | accuracy (%): 89.4 -> 90.3 | Helped |
| Use CoWA-JMDS instead of NRC | SFUDA · Office-Home closed-set, Avg | accuracy (%): 72.2 -> 72.5 | Helped |
| Use CoWA-JMDS instead of NRC | SFUDA · VisDA-2017 closed-set, Average | accuracy (%): 85.9 -> 86.9 | Helped |
| Add weight Mixup to CoWA-JMDS (ablation) | Office-31 closed-set adaptation · Office-31, Avg. | accuracy (%): 89.6 -> 90.3 | Helped |
| Add weight Mixup to CoWA-JMDS (ablation) | Office-Home closed-set adaptation · Office-Home, Avg | accuracy (%): 72.2 -> 72.5 | Helped |
| Add weight Mixup to CoWA-JMDS (ablation) | VisDA-2017 closed-set adaptation · VisDA-2017, Average | accuracy (%): 84.2 -> 86.9 | Helped |
| Add weight Mixup to CoWA-JMDS (ablation) | Open-set Office-Home adaptation · Office-Home open-set, Avg | accuracy (%): 74.2 -> 73.2 | Hurt |
| Use JMDS, combining LPG and MPPL, rather than MPPL alone | Confidence scoring with GMM pseudo-labels · Office-Home, Avg | AURC: 0.169 -> 0.151 | Helped |
| Use JMDS, combining LPG and MPPL, rather than LPG alone | Confidence scoring with GMM pseudo-labels · Office-Home, Avg | AURC: 0.168 -> 0.151 | Helped |
| Replace ordinary Mixup with weight Mixup | GMM pseudo-label adaptation · Office-31, Avg. | accuracy (%): 86.9 -> 90.3 | Helped |
Closest papers by meaning
- Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · similarity 0.89)
- 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 0.86)
- Proxy Denoising for Source-Free Domain Adaptation (ICLR 2025 · similarity 0.85)
- Can We Evaluate Domain Adaptation Models Without Target-Domain Labels? (ICLR 2024 · similarity 0.85)
- Distinguish Then Exploit: Source-free Open Set Domain Adaptation via Weight Barcode Estimation and Sparse Label Assignment (CVPR 2025 · 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.