s-chh / Pytorch-cGAN-conditional-GAN

Pytorch implementation of conditional generative adversarial network (cGAN) using DCGAN architecture for generating 32x32 images of MNIST, SVHN, FashionMNIST, and USPS datasets.

Pytorch-cGAN-conditional-GAN

Pytorch implementation of conditional generative adversarial network (cGAN) using DCGAN architecture for generating 32x32 images of MNIST, SVHN, FashionMNIST, and USPS datasets.

Change the DB variable to change the dataset.

For using the saved model to generate images, set LOAD_MODEL to True and EPOCHS to 0.

Generated Samples

MNIST

SVHN

FashionMNIST

USPS