Research map / Label-efficient and robust learning / Domain adaptation and generalization
Attracting and Dispersing: A Simple Approach for Source-free 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 |
|---|---|---|---|
| Apply AaD instead of NRC | Source-free domain adaptation with ResNet-101 · VisDA-C | Per-class accuracy: 85.9 -> 88.0 | Helped |
| Apply AaD instead of A²Net | Source-free domain adaptation with ResNet-50 · Office-Home Avg | accuracy: 72.8 -> 72.7 | Hurt |
| Apply AaD instead of HCL | Source-free domain adaptation with ResNet-50 · Office-31 Avg | accuracy: 89.8 -> 89.9 | Helped |
| Use AaD rather than SHOT | Source-free open-set DA on Office-Home · Office-Home open-set Avg | HOS: 45.6 -> 63.6 | Helped |
| Use 3 rather than 1 nearest neighbor (ablation) | AaD on Office-31 · Office-31 Avg | accuracy: 89.1 -> 89.9 | Helped |
| Use 3 rather than 1 nearest neighbor (ablation) | AaD on Office-Home · Office-Home Avg | accuracy: 72.2 -> 72.7 | Helped |
| Use 5 rather than 3 nearest neighbors (ablation) | AaD on VisDA-C · VisDA-C | Per-class accuracy: 86.7 -> 88.0 | Helped |
| Use 6 or 7 rather than 5 nearest neighbors (ablation) | AaD on VisDA-C · VisDA-C | Per-class accuracy: 88.0 -> 88.0 | No clear effect |
| Increase β from 0 to 2 (ablation) | AaD on Office-31 · Office-31 | accuracy: 88.0 -> 89.9 | Helped |
| Increase β from 0 to 1 (ablation) | AaD on Office-Home · Office-Home | accuracy: 72.7 -> 70.6 | Hurt |
| Increase β from 0 to 7 (ablation) | AaD on VisDA-C · VisDA-C | accuracy: 77.5 -> 88.1 | Helped |
| Use a 10% target-feature memory bank instead of the full bank (ablation) | AaD on VisDA-C · VisDA-C | Per-class accuracy: 88.0 -> 87.6 | Hurt |
Closest papers by meaning
- Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · similarity 0.90)
- SF(DA)$^2$: Source-free Domain Adaptation Through the Lens of Data Augmentation (ICLR 2024 · similarity 0.89)
- Dynamic Target Distribution Estimation for Source-Free Open-Set Domain Adaptation (AAAI 2025 · similarity 0.89)
- Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.89)
- Reviewing the Forgotten Classes for Domain Adaptation of Black-Box Predictors (AAAI 2024 · similarity 0.88)
- Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm (AAAI 2025 · similarity 0.88)
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.