Conditional Generative Adversarial Network for Devanagari Handwritten Character Generation

In the deep learning era, the model requires more variations in the large dataset for better understanding. Generative Adversarial Networks (GANs) are often used to learn to generate various images. The design of GAN consists of a collection of neural networks used in an unsupervised way. In GAN, many kinds of arrangements of neural networks are possible for various applications in many areas of computer vision and machine learning. Conditional Generative Adversarial Network (CGAN) is one of them helpful in the generation of the definite type of images. This work illustrates the use of CGAN for generating specific Devanagari handwritten characters.