Research map / Label-efficient and robust learning / Domain adaptation and generalization
SFHarmony: Source Free Domain Adaptation for Distributed Neuroimaging Analysis
Nicola K Dinsdale, Mark Jenkinson, Ana IL Namburete
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 |
|---|---|---|---|
| use SFHarmony GMM alignment instead of other SFDA methods at small batchsize | SFHarmony · OrganAMNIST (batchsize 5, simulated shifts) | average accuracy: 83.10 -> 86.22 | Improved |
| use 2 GMM components instead of 1 (ablation) | SFHarmony · OrganAMNIST (batchsize 5, simulated shifts) | average accuracy: 85.47 -> 86.22 | Improved |
| use 3 GMM components instead of 2 (ablation) | SFHarmony · OrganAMNIST (batchsize 5, simulated shifts) | average accuracy: 86.22 -> 86.21 | No clear effect |
| fit GMM with EM instead of direct mean and variance fit (ablation) | SFHarmony · OrganAMNIST (batchsize 5, simulated shifts) | average accuracy: 77.26 -> 85.47 | Improved |
| add EM batch memory across training iterations (ablation) | SFHarmony · OrganAMNIST (batchsize 5, simulated shifts) | average accuracy: 80.25 -> 85.47 | Improved |
| use SFHarmony instead of entropy-minimisation SFDA for segmentation | SFHarmony · CC359 brain extraction (batchsize 50) | average Dice: 0.849 -> 0.970 | Improved |
| use SFHarmony instead of Direct Fit single-Gaussian fit (ablation) | SFHarmony · ABIDE tissue segmentation (batchsize 50) | average Dice: 0.803 -> 0.832 | Improved |
| use SFHarmony instead of FADA (which has source data) for age regression | SFHarmony · ABIDE age prediction (batchsize 8) | average MAE: 3.42 -> 3.72 | Hurt |
| vary learning rate from 1e-7 to 1e-4 (ablation) | SFHarmony · OrganAMNIST | average accuracy | No clear effect |
Closest papers by meaning
- Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · similarity 0.87)
- BlindHarmony: "Blind" Harmonization for MR Images via Flow Model (ICCV 2023 · similarity 0.87)
- Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.86)
- Supportive Negatives Spectral Augmentation for Source-Free Cross-Domain Segmentation (AAAI 2025 · similarity 0.86)
- AIF-SFDA: Autonomous Information Filter Driven Source-Free Domain Adaptation for Medical Image Segmentation (AAAI 2025 · similarity 0.85)
- Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm (AAAI 2025 · similarity 0.85)
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.
