Deep learning architectures such as deep neural networks and recurrent neural networks are applied to fields including computer vision, speech recognition, natural language processing, , drug design, medical image analysis, reinforcement learning etc.
One of the interesting application of deep learning is
Generative Adversarial Networks (GANs) which aim at learning the true data distribution of the training set so as to generate new data points. Since the day it was introduced in 2014 it has been used in various applications from Image Synthesis to synthesizing DNA sequences.
This project aims to provide flexible and extensible implementation of
Least Squares Generative Adversarial Networks(LSGANs),
Bidirectional GAN(BIGAN) and
Stacked Generative Adversarial Network (StackGAN). In addition it also aims at providing techniques to measure GAN performance like
Inception Score(IS) and
Fréchet Inception Distance(FID).