AI Paper Atlas

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

ECCV 2022 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Source-free domain adaptation · 73 citations (OpenAlex)

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 changedWhereMeasuredEffect
replace monocentric prototype pseudo labels with class-balanced sampling (BP) (ablation)SHOT · Office-Homeaccuracy (%): 71.5 -> 72.0Improved
replace monocentric prototype pseudo labels with class-balanced sampling (BP) (ablation)SHOT · VisDA-Caccuracy (%): 82.9 -> 83.8Improved
replace monocentric prototype pseudo labels with class-balanced sampling (BP) (ablation)SHOT · PointDA-10accuracy (%): 53.1 -> 55.0Improved
add intra-class multicentric prototypes to BP (BMP) (ablation)SHOT w/ BP · Office-Homeaccuracy (%): 72.0 -> 72.5Improved
add intra-class multicentric prototypes to BP (BMP) (ablation)SHOT w/ BP · VisDA-Caccuracy (%): 83.8 -> 84.7Improved
add intra-class multicentric prototypes to BP (BMP) (ablation)SHOT w/ BP · PointDA-10accuracy (%): 55.0 -> 56.4Improved
add dynamic EMA pseudo labels to BMP (ablation)SHOT w/ BMP · Office-Homeaccuracy (%): 72.5 -> 72.5No clear effect
add dynamic EMA pseudo labels to BMP (ablation)SHOT w/ BMP · VisDA-Caccuracy (%): 84.7 -> 85.7Improved
add dynamic EMA pseudo labels to BMP (ablation)SHOT w/ BMP · PointDA-10accuracy (%): 56.4 -> 57.0Improved
add BMD strategy to existing SFDA methodsSHOT · VisDA-Caccuracy (%): 82.9 -> 85.8Improved
add BMD strategy to existing SFDA methodsSHOT++ · VisDA-C truck classaccuracy (%): 28.8 -> 45.9Improved
add BMD strategy to NRCNRC · VisDA-C truck classaccuracy (%): 59.7 -> 55.6Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy