hparik11 / face-generation

Using generative adversarial networks (GAN) to generate new images of faces from celebrity images.

Face Generation DCGAN

Using deep learning DCGAN neural network to generate realistic faces from celebrity images.

Overview

Using a Deep Convolutional Generative Adversarial Network (DCGAN) to generates new images of faces. The dataset that's used is the CelebFaces Atrribute Dataset (CelebA), which contains over 200,000 celebrity faces with annotations.

Tensorflow implementation of Deep Convolutional Generative Adversarial Networks which is a stabilize Generative Adversarial Networks.

Files

Jupyter Notebook

Hyperparameters

Hyperparameter	Number
Epochs	1
Batch size	128
Learning rate	0.0005
Z dimension	100
Beta 1	0.5

Results

celebA

After 10 iterations:

After 20 iterations:

MNIST

After 10 iterations:

After 20 iterations:

Related works

• DCGAN-tensorflow
• DiscoGAN-pytorch
• simulated-unsupervised-tensorflow