README.md

PyTorch implementation of Generative Adversarial Networks (GAN)

Implemented models:

• GAN
• DCGAN
• Conditional DCGAN

Analyzed Datasets:

• MNIST
• CelebA

MNIST

GAN

Learning curves

Generated examples

examples.mp4

DCGAN

Learning curves

Generated examples

video.mp4

Conditional DCGAN

Learning curves

Examples (each row is conditioned with specific digit)

examples.mp4

CelebA

GAN

Learning curves

Generated examples

video.mp4

DCGAN

Learning curves

Generated examples

video.mp4

Interpolation of latent variable

For each latent dimension latent_dim a new z noise was sampled from gaussian distribution and z[latent_dim] was then interpolated between -3.5 and 3.5