Implementations(chiefly in Keras w/Tensorflow backend) of various Deep Generative Models(e.g. Variational Autoencoders, RBMs, and Generative Adversarial Networks)
-
Variational_Autoencoder.py employs a Variational Autoencoder (Kingma, Welling 2013) to reconstruct MNIST handwritten digits.
-
Deep_Convolutional_GAN.py uses a Deep Convolutional Generative Adversarial Network (Radford et. al, 2015) to reconstruct MNIST handwritten digits.