Medical Image Classification on FractureMNIST3D
0.774AUCG-LoG bi-filtration (MLP)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| G-LoG bi-filtration (MLP)sigma=1.52026.02 | 0.774 | 0.588 | |
| G-LoG bi-filtration (MLP)sigma=12026.02 | 0.753 | 0.579 | |
| ResNet-50+ACS2026.02 | 0.75 | 0.517 | |
| BSDA# of trainable params=33M2026.02 | 0.731 | 0.569 | |
| ResNet-50+3D2026.02 | 0.725 | 0.494 | |
| CdTransformer + CcCL2026.02 | 0.724 | 0.529 | |
| AC-Base# of trainable params=427M2026.02 | 0.719 | 0.608 | |
| AC-Small# of trainable params=216M2026.02 | 0.717 | 0.563 | |
| ResNet-18+ACS2026.02 | 0.714 | 0.497 | |
| ResNet-18+3D2026.02 | 0.712 | 0.508 | |
| Official# of trainable params=33M2026.02 | 0.712 | 0.508 | |
| G-LoG bi-filtration (MLP)sigma=0.52026.02 | 0.704 | 0.579 | |
| G-LoG bi-filtration (MLP)sigma=02026.02 | 0.697 | 0.554 | |
| LMTTM-VMI# of trainable params=21M2026.02 | 0.691 | 0.562 | |
| AC-Tiny# of trainable params=9.5M2026.02 | 0.686 | 0.55 | |
| ViViT-M + R-LLM# of trainable params=295M2026.02 | 0.669 | 0.563 | |
| Topo-Med (MLP)2026.02 | 0.653 | 0.48 | |
| AutoKeras2026.02 | 0.642 | 0.458 | |
| auto-sklearn2026.02 | 0.628 | 0.453 | |
| ResNet-18+2.5D2026.02 | 0.587 | 0.451 | |
| ResNet-50+2.5D2026.02 | 0.552 | 0.397 | |
| Medformer2026.02 | 0.536 | 0.427 |