GAN-based denoising methods usually leverage an additional classification network, i.e., discriminator, to learn the most discriminate difference between the denoised and normal-dose images and, hence ...