Research map / Label-efficient and robust learning / Domain adaptation and generalization
MetaTeacher: Coordinating Multi-Model Domain Adaptation for Medical Image Classification
Zhenbin Wang, Mao Ye, Xiatian Zhu, Liuhan Peng, Liang Tian, Yingying Zhu
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 coordinating weight learning (vs averaging teachers) (ablation) | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Open-i | AUROC (average): 90.76 -> 92.84 | Improved |
| add bilevel optimization of teachers and student (ablation) | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Open-i | AUROC (average): 89.32 -> 92.84 | Improved |
| add coordinating weight learning (vs averaging teachers) (ablation) | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Google-Health-CXR | AUROC (average): 80.35 -> 82.69 | Improved |
| add bilevel optimization of teachers and student (ablation) | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Google-Health-CXR | AUROC (average): 79.08 -> 82.69 | Improved |
| add coordinating weight learning (vs averaging teachers) (ablation) | MetaTeacher · CheXpert+MIMIC-CXR -> NIH-CXR14 | AUROC (average): 76.16 -> 77.74 | Improved |
| add bilevel optimization of teachers and student (ablation) | MetaTeacher · CheXpert+MIMIC-CXR -> NIH-CXR14 | AUROC (average): 75.16 -> 77.74 | Improved |
| add coordinating weight learning (vs averaging teachers) (ablation) | MetaTeacher · NIH-CXR14+CheXpert -> Open-i | AUROC (average): 90.72 -> 91.98 | Improved |
| add bilevel optimization of teachers and student (ablation) | MetaTeacher · NIH-CXR14+CheXpert -> Open-i | AUROC (average): 88.40 -> 91.98 | Improved |
| increase labeled target data from 50 to 200 | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Open-i | AUROC (average): 91.09 -> 92.84 | Improved |
| increase labeled target data from 200 to 300 | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Open-i | AUROC (average): 92.84 -> 92.97 | No clear effect |
| set beta to 0 (no mapping regularization) (ablation) | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Open-i | AUROC (average): 92.84 -> 92.49 | Hurt |
| add divergence loss (gamma 0 -> 0.01) (ablation) | MetaTeacher · NIH-CXR14+CheXpert+MIMIC-CXR -> Open-i | AUROC (average): 92.29 -> 92.84 | Improved |
Closest papers by meaning
- Taxonomy Adaptive Cross-Domain Adaptation in Medical Imaging via Optimization Trajectory Distillation (ICCV 2023 · similarity 0.88)
- Steady Progress Beats Stagnation: Mutual Aid of Foundation and Conventional Models in Mixed Domain Semi-Supervised Medical Image Segmentation (CVPR 2025 · similarity 0.86)
- Constructing and Exploring Intermediate Domains in Mixed Domain Semi-supervised Medical Image Segmentation (CVPR 2024 · similarity 0.85)
- Leveraging Anatomical Consistency for Multi-Object Detection in Ultrasound Images via Source-free Unsupervised Domain Adaptation (AAAI 2025 · similarity 0.85)
- Systematic Comparison of Semi-supervised and Self-supervised Learning for Medical Image Classification (CVPR 2024 · similarity 0.84)
- Versatile Medical Image Segmentation Learned from Multi-Source Datasets via Model Self-Disambiguation (CVPR 2024 · similarity 0.84)
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.
