Research map / Label-efficient and robust learning / Domain adaptation and generalization
BMD: A General Class-Balanced Multicentric Dynamic Prototype Strategy for Source-Free Domain Adaptation
Sanqing Qu, Guang Chen, Jing Zhang, Zhijun Li, Wei He, Dacheng Tao
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 |
|---|---|---|---|
| replace monocentric prototype pseudo labels with class-balanced sampling (BP) (ablation) | SHOT · Office-Home | accuracy (%): 71.5 -> 72.0 | Improved |
| replace monocentric prototype pseudo labels with class-balanced sampling (BP) (ablation) | SHOT · VisDA-C | accuracy (%): 82.9 -> 83.8 | Improved |
| replace monocentric prototype pseudo labels with class-balanced sampling (BP) (ablation) | SHOT · PointDA-10 | accuracy (%): 53.1 -> 55.0 | Improved |
| add intra-class multicentric prototypes to BP (BMP) (ablation) | SHOT w/ BP · Office-Home | accuracy (%): 72.0 -> 72.5 | Improved |
| add intra-class multicentric prototypes to BP (BMP) (ablation) | SHOT w/ BP · VisDA-C | accuracy (%): 83.8 -> 84.7 | Improved |
| add intra-class multicentric prototypes to BP (BMP) (ablation) | SHOT w/ BP · PointDA-10 | accuracy (%): 55.0 -> 56.4 | Improved |
| add dynamic EMA pseudo labels to BMP (ablation) | SHOT w/ BMP · Office-Home | accuracy (%): 72.5 -> 72.5 | No clear effect |
| add dynamic EMA pseudo labels to BMP (ablation) | SHOT w/ BMP · VisDA-C | accuracy (%): 84.7 -> 85.7 | Improved |
| add dynamic EMA pseudo labels to BMP (ablation) | SHOT w/ BMP · PointDA-10 | accuracy (%): 56.4 -> 57.0 | Improved |
| add BMD strategy to existing SFDA methods | SHOT · VisDA-C | accuracy (%): 82.9 -> 85.8 | Improved |
| add BMD strategy to existing SFDA methods | SHOT++ · VisDA-C truck class | accuracy (%): 28.8 -> 45.9 | Improved |
| add BMD strategy to NRC | NRC · VisDA-C truck class | accuracy (%): 59.7 -> 55.6 | Hurt |
Closest papers by meaning
- Divide and Contrast: Source-free Domain Adaptation via Adaptive Contrastive Learning (NeurIPS 2022 · similarity 0.92)
- Unraveling the Mysteries of Label Noise in Source-Free Domain Adaptation: Theory and Practice (TPAMI 2025 · similarity 0.92)
- Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.91)
- Generalized Source-Free Domain Adaptation (ICCV 2021 · similarity 0.91)
- Dynamic Target Distribution Estimation for Source-Free Open-Set Domain Adaptation (AAAI 2025 · similarity 0.90)
- Diffusion-Driven Progressive Target Manipulation for Source-Free Domain Adaptation (NeurIPS 2025 · 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.
