iMTE

Super-resolution GAN (SRGAN)Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial NetworkLedig, Christian, et al. "Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network." CVPR. Vol. 2. No. 3. 2017. 간략하게 설명하자면, 기존의 super-resolution 문제에 대해서 단순히 MSE(Mean squared error)를 loss function(objective function)으로 정해놓고 optimization을 해본 결과, high peak sign..