Human Face Image Generation from Textual Descriptions

Listify Chrome Extension

This project focuses on the innovative application of Generative Adversarial Networks (GANs), specifically a variant known as Text-Guided Diverse Face Image Generation (Tedi-GAN). This technology was fine-tuned using Tensorflow and Pytorch to achieve detailed and accurate face generation based on textual descriptions provided by users.

Project Overview:

  • Advanced GAN Model: Utilized Tedi-GAN, a sophisticated model that interprets textual descriptions to generate diverse human face images. The model’s performance was enhanced by fine-tuning adjustments, improving accuracy by 68%.
  • Web Interface: Developed an intuitive web platform that allows users to enter descriptions of individuals. The site uses the Tedi-GAN model to generate corresponding high-definition images of the described faces.
  • Collaborative Development: Worked alongside two other developers to architect the front-end of the website using React, ensuring a user-friendly and responsive design.

This project not only demonstrates advanced capabilities in machine learning and image generation but also provides a practical application that users can interact with directly to see the power of AI in creative tasks.