Table 2 Quantitative results associated with different network outputs for Figs. 6 and 7.
From: A generative adversarial network with “zero-shot” learning for positron image denoising
Fig. 6 | Fig.7 | |||
|---|---|---|---|---|
PSNR | SSIM | PSNR | SSIM | |
DnCNN | 27.872 | 0.804 | 27.954 | 0.704 |
GAN | 28.326 | 0.635 | 27.898 | 0.683 |
CNN-GAN | 27.928 | 0.856 | 30.127 | 0.742 |
Zero-GAN | 29.061 | 0.892 | 29.568 | 0.837 |
Our method | 29.342 | 0.897 | 31.783 | 0.925 |
