Research map / Label-efficient and robust learning / Domain adaptation and generalization
Trust Your Good Friends: Source-Free Domain Adaptation by Reciprocal Neighborhood Clustering
Shiqi Yang, Yaxing Wang, Joost van de Weijer, Luis Herranz, Shangling Jui, Jian Jun Yang
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 neighbor supervision L_N to diversity loss (ablation) | NRC · Office-Home | avg accuracy: 62.1 -> 69.1 | Improved |
| add affinity values for reciprocal neighbors (ablation) | NRC · Office-Home | avg accuracy: 69.1 -> 71.1 | Improved |
| add expanded neighborhoods together with affinity (ablation) | NRC · Office-Home | avg accuracy: 71.1 -> 72.2 | Improved |
| add expanded neighbors without affinity (ablation) | NRC · Office-Home | avg accuracy: 69.1 -> 65.2 | Hurt |
| add expanded neighbors without affinity (ablation) | NRC · VisDA | per-class accuracy: 81.5 -> 61.2 | Hurt |
| add affinity to expanded neighbors (ablation) | NRC · VisDA | per-class accuracy: 61.2 -> 85.9 | Improved |
| add outlier exclusion density loss L_D (NRC++) (ablation) | NRC · VisDA | per-class accuracy: 85.9 -> 88.1 | Improved |
| add outlier exclusion density loss L_D (NRC++) (ablation) | NRC · Office-Home | avg accuracy: 72.2 -> 72.5 | Improved |
| add outlier exclusion density loss L_D (NRC++) (ablation) | NRC · Office-Home partial-set DA | avg accuracy: 79.5 -> 79.6 | No clear effect |
| use expanded neighbors instead of larger K nearest neighbors (ablation) | NRC · VisDA | per-class accuracy: 84.0 -> 85.9 | Improved |
| use expanded neighbors instead of larger K nearest neighbors (ablation) | NRC · Office-Home | avg accuracy: 69.5 -> 72.2 | Improved |
| remove duplication in expanded neighbors (ablation) | NRC · VisDA | per-class accuracy: 85.9 -> 82.0 | Hurt |
Closest papers by meaning
- Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation (NeurIPS 2021 · similarity 0.97)
- Source-Free Domain Adaptation via Distribution Estimation (CVPR 2022 · similarity 0.92)
- SF(DA)$^2$: Source-free Domain Adaptation Through the Lens of Data Augmentation (ICLR 2024 · similarity 0.91)
- Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.89)
- Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · similarity 0.89)
- Dynamic Target Distribution Estimation for Source-Free Open-Set Domain Adaptation (AAAI 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.
