Skip to content

This project implements a Text to Image Synthesis system using Deep Convolutional Generative Adversarial Networks (DCGAN). The goal is to generate photo-realistic images from textual descriptions, demonstrating the power of deep learning in creative applications.

Notifications You must be signed in to change notification settings

Kaifansar1/Text-to-Image-Synthesis-using-Deep-Learning-Python-PyTorch

Repository files navigation

Text To Image Synthesis Using DCGAN

687474703a2f2f692e696d6775722e636f6d2f644e6c32486b5a2e6a7067

Image Source : Generative Adversarial Text-to-Image Synthesis Paper link

Details of the Project Repository:

  • TextToImageSynthesizerDCGANFinal1.0.ipynb is the final notebook and the one that needs to be run and evaluated.
  • The Notebook contains all the instructions to run the notebook.
  • checkpoints: This folder consists of the final generator and discriminator model after 100 Epochs of training and can be used as a pretrained model.
  • results: This folder consists of images generated by dcgan on training, validation and test split.
  • epoch_loss.csv: This consists of the generator and discriminator losses for the number of training epochs.

About

This project implements a Text to Image Synthesis system using Deep Convolutional Generative Adversarial Networks (DCGAN). The goal is to generate photo-realistic images from textual descriptions, demonstrating the power of deep learning in creative applications.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published