latentcollection/glyph-gan

Pytorch based deep convolutional GAN trained on glyphs

GylphGAN

Pytorch based deep convolutional GAN trained on glyphs. Based on the original architecture by Ritchie Vink, adapted by Aleksi Halttunen, and implemented by Moritz Salla.

Make sure to place your dataset in a subdir from root:

data = torch.utils.data.DataLoader(ImageFolderEX(".", trans),
       batch_size=batch_size, shuffle=True,
       drop_last=True, num_workers=2)

.
└── 1
    ├── img1.jpg
    ├── img2.jpg
    └── img3.jpg