Research on GAN-based image enhancement method

To address the topic of fuzzy face image enhancement, this paper proposes a method based on GAN network, the overall structure of CycleGAN, in the network we modify the traditional CycleGAN structure, add the residual network layer, randomly selected images in FFHQ dataset and face normalization for testing, through a series of evaluation indexes to carry out quantitative assessment. In addition, a qualitative and quantitative comparison is made with related research methods in recent years to verify the advantages of this method in face image enhancement.