Generating Synthetic Images with DCGANs(Deep Convolutional GAN) in Keras

By Vimal Kumar Verma

In this Packet, i am going to build and train a Deep Convolutional GAN (DCGAN) with Keras to generate images of digits.I am going to use mnist dataset.

In this project I’m going to build a specialized neural network architecture called a Generative Adversarial Network (GAN).Two main part of any GAN network that is Generater and Discriminater. We keep the generator constant during the discriminator training phase. As discriminator training tries to figure out how to distinguish real data from fake, it has to learn how to recognize the generator's flaws. ... Similarly, we keep the discriminator constant during the generator training phase

                                    

This project is written in Python and uses the following libraries and frameworks:

1. tensorflow-2.3.0

2. Matplotlib-3.1.0

3. NumPy-1.16.4

4.tqdm-4.32.1

And at the end of this packet you will get a gif format of augumented images