https://nayopu.github.io/eyes2mouth/
- Download celeba dataset
$ ./download_celeba/sh
- Cropping faces
$ python gen_dataset.py
- Start training
$ python main.py --phase train --dataset_name cropped_128 --batch_size 10 --fine_size 128 --lr 0.00002
- Start testing
$ python main.py --phase test --dataset_name cropped_128 --batch_size 10 --fine_size 128