Digit Generation using Conditional GAN (C-GAN)

• Concepts used:
  • Generator
  • Discriminator (Classifier)
  • Batch Normalization
  • Transpose Convolution
• Tech Stack:
• Python and PyTorch
• GitHub: Project Link

Trained a model on MNIST dataset using Conditional Generative Adversarial Neural Network(GAN) in Python using PyTorch framework and our model was able to produce good results than the vanilla GAN. It also overcome the problem that the DCGAN had and is now able to produce images based on the conditions such as specific class if we want to give like shirts etc. It got possible now due to the class label(y) which is fed to both Generator and Discriminator.