Super-FAN
Super-FAN copied to clipboard
jzijin
The PyTorch implement of the paper "Super-FAN: Integrated facial landmark localization and super-resolution of real-world low resolution faces in arbitrary poses with GANs"
Super-FAN Pytorch
The PyTorch implement of Super-FAN. Super-FAN: Integrated facial landmark localization and super-resolution of real-world low resolution faces in arbitrary poses with GANs Based on SRGAN, I altered the code by adopting Super-FAN network structure.
Prerequisites
- Python 3.6
- Pytorch 1.0 or newer
Dataset
Use the same dataset of FSRNet.Change the option in Train.py to set the dataset's directory. I am using CelebAHQ-MASK as the training set. The GroundTruth is generated by zllrunning/face-parsing.PyTorch(https://github.com/zllrunning/face-parsing.PyTorch) with pretrained model.
Dataset Link: https://pan.baidu.com/s/1HEECUyKI5GOSrd7NPlm-ow 密码:z2ud
Train and Test
I haved merge the Super-FAN in to the mmsr.I use the mmsr hierarchical to pretrain Super-FAN's Generator.the 3000 of the CelebAHQ-MASK used for train with 212000 iter.
Please read the mmsr for training and testing.
Result
The pretrain model result.
The left to right is bicubic interpolation image, high resolution image, SR image without GAN, SR image with GAN._
jzijin