Table 1 Algorithm comparison.
From: Discriminative Learning for Automatic Staging of Placental Maturity via Multi-layer Fisher Vector
Reference | Sample | Feature | Validation | Classifier | Result |
|---|---|---|---|---|---|
Linares16 | 59 | textural features such as co-occurrence matrices, Laws masks and neighborhood gray-tone difference matrice | Leave one out | KNN | Best accuracy: 60.71% |
Liu et al.15 | 200 | gray level statistical feature: mean, variance, distortions and kurtosis of gray scale. | 120 training, 80 testing | SVM | Recognition rate 92% |
Li et al.20 | 311 | Dense sampling, DAISY descriptor. | Random partition | SVM | mAP: 92.5%, Sensitivity: 99.6%, Accuracy: 87.4% |
Lei et al.21 | 443 | HarrisLaplace, LIOP descriptor | Random partition | SVM | Accuracy: 93.75%,Sensitivity: 98.04% Specificity: 93.75% |
Proposed | 443 | Visual descriptor such as DAISY, LIOP, SIFT, dense sampling | 10 fold cross-validation | SVM | AUC: 96.77%, Sensitivity: 98.04%, Specificity: 93.75% |
