Applied-AI-Lab-Deep-Learning-for-Computer-Vision

WorldQuant University

🏥 Medical Image Generation with GANs

🌟 Project Overview

The tasks involved in this project explored Generative Adversarial Networks (GANs) to generate synthetic medical images like X-rays and MRIs. 🏥 You'll build a custom GAN from scratch and also use a pre-trained GAN to create realistic images. 🎨 To make it interactive, you'll develop a Streamlit web app that allows users to generate medical images dynamically. 🌐 You'll also use Git and GitHub for version control and collaboration. 🗂️

Key Components:

1. Generative Adversarial Networks (GANs):

   • A custom GAN was designed and trained to generate realistic medical images.
   • A pre-trained GAN was employed to enhance efficiency in the creation of synthetic datasets.

2. Synthetic Data Utilization:

   • Realistic medical images, including X-rays and MRIs, were generated using the GAN.
   • These synthetic images were utilized to train and evaluate machine learning models.

3. Web App Development:

   • An interactive web application was created using Streamlit, providing users with the ability to generate and visualize medical images.

4. Version Control and Collaboration:

   • Code changes were tracked using Git, and the project was shared through GitHub repositories to ensure collaborative development and proper version control.

Skills and Knowledge Gained:

• 🤖 How to build and train a GAN from scratch
• 🖼️ How to generate images using a pre-trained GAN
• 🏥 How to train models with synthetic medical data
• 🌐 How to build a web app using Streamlit
• 🗂️ How to track and share code with Git & GitHub

This project is a great way to explore AI in healthcare and build hands-on experience with GANs! 🚀✨

Data & Models: https://drive.google.com/drive/folders/1rfj7ZpYV4rb1pOpncX6RfclRsDBrL9Pa?usp=sharing

---