Historical Image Dating on Historical color image dataset (test)
54.68AccuracyClassification + POE
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Classification + POEBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 54.68 | 0.67 | |
| CNNPORBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 50.12 | 0.82 | |
| Ranking + POEBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 50.01 | 0.66 | |
| ClassificationBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 48.94 | 0.89 | |
| GP-DNNORBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 46.6 | 0.76 | |
| D-Regression + POEBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 46.04 | 0.72 | |
| Palermo et al.Backbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 44.92 | 0.93 | |
| RankingBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 44.67 | 0.81 | |
| D-RegressionBackbone=VGG-16, Pre-training=ImageNet, Evaluation Protocol=ten-fold cross-validation2021.03 | 42.24 | 0.79 |