Overview / Usage

We are working on a novel method to convert a GUI screenshot to front-end code using Generative Adversarial Networks. We are using Policy Gradient (Reinforcement Learning) and Monte Carlo Tree Search to train the Generator to generate discrete markup tokens. Through our research, we want to contribute to the exciting revolution to create code that writes code

Methodology / Approach

Using a conditional GAN with variations of CNN-LSTM architectures for both Generator and Discriminator. Training the Generator using Policy Gradient (Reinforcement Learning) and Monte Carlo Tree Search.

Technologies Used

PyTorch, Torch, Intel Hardware CPU,