Research map / Label-efficient and robust learning / Domain adaptation and generalization
Uncertainty-Guided Source-Free Domain Adaptation
Subhankar Roy, Martin Trapp, Andrea Pilzer, Juho Kannala, Nicu Sebe, Elisa Ricci, Arno Solin
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 Laplace-approximation uncertainty-weighted entropy loss to SHOT-IM | SHOT-IM · Office-Home closed-set | avg accuracy (%): 70.5 -> 71.8 | Improved |
| weight entropy loss with MAP-model certainty instead of no weighting (ablation) | SHOT-IM · Office-Home closed-set | avg accuracy (%): 70.5 -> 71.2 | Improved |
| replace MAP weights with Laplace-approximation (posterior) weights (ablation) | SHOT-IM + entropy weighting · Office-Home closed-set | avg accuracy (%): 71.2 -> 71.8 | Improved |
| add uncertainty-guided weighting to SHOT-IM | SHOT-IM · Office-Home open-set | OS accuracy (%): 71.5 -> 73.5 | Improved |
| add uncertainty-guided weighting to SHOT-IM | SHOT-IM · VisDA-C | accuracy (%): 80.3 -> 81.2 | Improved |
| add uncertainty-guided weighting to SHOT-IM | SHOT-IM · DomainNet (avg) | accuracy (%): 21.98 -> 26.13 | Improved |
| add uncertainty-guided weighting to SHOT-IM, source quickdraw | SHOT-IM · DomainNet (source quickdraw) | accuracy (%): 10.76 -> 10.44 | Hurt |
| add uncertainty-guided weighting to SHOT-IM | SHOT-IM · CIFAR9 to STL9 | accuracy (%): 75.69 -> 77.04 | Improved |
| add uncertainty-guided weighting to SHOT-IM | SHOT-IM · Office31 (saturated) | avg accuracy (%): 87.3 -> 88.6 | Improved |
| add nearest-centroid pseudo-labelling of SHOT to U-SFAN | U-SFAN · Office-Home closed-set | avg accuracy (%): 71.8 -> 71.9 | No clear effect |
| add nearest-centroid pseudo-labelling of SHOT to U-SFAN | U-SFAN · VisDA-C | accuracy (%): 81.2 -> 82.7 | Improved |
| add nearest-centroid pseudo-labelling of SHOT to U-SFAN | U-SFAN · Office31 | avg accuracy (%): 88.6 -> 88.8 | No clear effect |
Closest papers by meaning
- Revisiting Source-Free Domain Adaptation: a New Perspective via Uncertainty Control (ICLR 2025 · similarity 0.94)
- Rethinking the Role of Pre-Trained Networks in Source-Free Domain Adaptation (ICCV 2023 · similarity 0.92)
- Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · similarity 0.91)
- When Source-Free Domain Adaptation Meets Learning with Noisy Labels (ICLR 2023 · similarity 0.91)
- Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.91)
- Revisiting the Domain Shift and Sample Uncertainty in Multi-source Active Domain Transfer (CVPR 2024 · similarity 0.90)
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.
