VQ-VAE-2 is a image synthesis model based on Variational. Autoencoders. It produces images that are high quality, comparable (FID/Inception).
確定! 回上一頁