Image Classification on ALL-IDB2
98.65AccuracyBayesian CNN + Augmentation
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Bayesian CNN + AugmentationMethodology=Bayesian CNN + Augmentation2026.05 | 98.65 | — | |
| PRISM (ET+SVM+LogReg)Computational Paradigm=Multiscale Zonal Extraction + Stacking2026.05 | 98.46 | 0.9698 | |
| CNN (ResNet + GA)Methodology=CNN (ResNet + GA)2026.05 | 98.46 | — | |
| Hybrid CNN Ensemble (MobileNet + ShuffleNet)Methodology=Hybrid CNN Ensemble (MobileNet + ShuffleNet)2026.05 | 98.46 | — | |
| PRISMMethodology=Zonal Stacking Ensemble2026.05 | 98.46 | — | |
| EfficientNet-B0Computational Paradigm=CNN2026.05 | 98.08 | 0.963 | |
| DenseNet-121Computational Paradigm=CNN2026.05 | 97.69 | 0.9557 | |
| VGG16Computational Paradigm=CNN2026.05 | 97.69 | 0.955 | |
| Handcrafted (SIFT+SURF)Methodology=Handcrafted (SIFT+SURF)2026.05 | 97.22 | — | |
| CNN (MobileNet + ResNet)Methodology=CNN (MobileNet + ResNet)2026.05 | 97.18 | — | |
| ResNet-18Computational Paradigm=CNN2026.05 | 96.54 | 0.9333 | |
| CNN (VGG + SESSA)Methodology=CNN (VGG + SESSA)2026.05 | 96.11 | — | |
| Handcrafted (KNN)Methodology=Handcrafted (KNN)2026.05 | 95.67 | — | |
| Swin Transformer (Swin-T)Computational Paradigm=Vision Transformer (Self-Attention)2026.05 | 91.92 | 0.8597 | |
| Handcrafted (EOF+MLP)Methodology=Handcrafted (EOF+MLP)2026.05 | 91.8 | — | |
| Hybrid CNN (ResNet + GA)Methodology=Hybrid CNN (ResNet + GA)2026.05 | 85 | — |