Research Article
A Computationally Virtual Histological Staining Method to Ovarian Cancer Tissue by Deep Generative Adversarial Networks
Table 3
Comparison of image quality using different loss functions and generator network structures.
| Network | Loss function | FID | IS |
| UNet | | 57.6092 | 1.3405 | | 54.1733 | 1.4687 | | 56.1733 | 1.4687 | | 59.4684 | 1.4720 | | 58.1196 | 1.4720 | | 56.1639 | 1.4432 |
| | | 54.1436 | 1.3928 | | 50.8299 | 1.4256 | | 52.4708 | 1.3865 | | 51.3790 | 1.3907 | | 55.2754 | 1.4852 | | 49.3387 | 1.4073 |
| PFFN (ours) | | 54.8384 | 1.3835 | | 49.1167 | 1.3903 | | 49.6818 | 1.4124 | | 49.3575 | 1.3238 | | 47.0977 | 1.3505 | | 48.8730 | 1.2158 |
|
|