The training process is divided into two stages. First, we train a PSNR-oriented model with the L1 loss. The obtained model is named by Real-ESRNet. We then use the trained PSNR-oriented model as an initialization of the generator, and train the Real-ESRGAN with a combination of L1 loss, perceptual loss and GAN loss


생성자의 초기화로 훈련된 PSNR-oriented model을 사용한다라는게 뭐 소리임?

애초에 PSNR-oriented model이라는게 뭐야?