Setup on | Originals | Fakes #1 | Fakes #1 | Fakes #2 | Fakes # 2 |
---|
(\(P_{miss}\)) | White (\(P_{fa}\)) | Gray (\(P_{fa}\)) | White (\(P_{fa}\)) | Gray (\(P_{fa}\)) |
---|
Fakes #1 white | 0 | 0 | 0.14 | 0 | 0 |
Fakes #1 gray | 0 | 0 | 0 | 0 | 0 |
Fakes #2 white | 0 | 99.43 | 100 | 0 | 0 |
Fakes # 2 gray | 0 | 99.29 | 99.86 | 0 | 0 |
- aPresented binary classification is close to the multi-class classification scenario with 2 classes considered in Section 3.2.1. The difference in the obtained results is related to the presence of all types of fakes during the training in case of multi-class setup and randomly chosen training data