DoB24
/

fundus-9model-benchmark

Image Classification

medical-imaging

retinal-disease

cross-validation

Model card Files Files and versions

fundus-9model-benchmark

4.11 GB

Ctrl+K

Ctrl+K

1 contributor

History: 76 commits

DoB24's picture

Add pooled-protocol provenance scripts (regenerate summary, 9-model McNemar, per-class, ensemble from fold predictions)

63fc39b verified about 2 months ago

analysis
Pooled five-fold consistency: foundation models pooled like CNN/CLIP; 9-model McNemar (36 pairs); per-class pooled; pooled ensemble report about 2 months ago
code
Add pooled-protocol provenance scripts (regenerate summary, 9-model McNemar, per-class, ensemble from fold predictions) about 2 months ago
dataset
Add dataset class distribution + sample grid 2 months ago
gradcam
gradcam v2: gradcam_vgg19.png 2 months ago
kfold
Pooled five-fold consistency: foundation models pooled like CNN/CLIP; 9-model McNemar (36 pairs); per-class pooled; pooled ensemble report about 2 months ago
results
Add 9-model fundus benchmark: weights + results + splits + code + README 2 months ago
splits
Add 9-model fundus benchmark: weights + results + splits + code + README 2 months ago
weights
Add 9-model fundus benchmark: weights + results + splits + code + README 2 months ago
.gitattributes

3.78 kB
gradcam v2: gradcam_swin_b.png 2 months ago
CITATION.cff

551 Bytes
Add CITATION.cff 2 months ago
README.md

19.2 kB
README v2: authoritative 5-fold CV table (§5.1) + single-run baseline (§5.2), verified metrics incl. kappa/brier/F1-CI 2 months ago