This repository contains a Generative Adversarial Network (GAN) implementation for generating synthetic breast cancer images using PyTorch. The GAN is trained on a dataset of breast cancer images.
- PyTorch Implementation: Utilizes the power of PyTorch for efficient deep learning model development.
- GAN Architecture: Implements a Generative Adversarial Network with both generator and discriminator components.
- Breast Cancer Image Synthesis: Generates synthetic breast cancer images to augment existing datasets.
Take a look at the final example of the generated image:
The generator and discriminator loss plots are also available:
Additionally, the saved models are G.pth (Generator) and D.pth (Discriminator).
- Data Preparation: Ensure your breast cancer image dataset is appropriately organized.
- Training: Run the entire notebook to train the GAN on the dataset.
- Inference: Generate synthetic images using the trained GAN model.
Feel free to explore and contribute to further advancements in synthetic image generation for breast cancer research.
- PyTorch
- The dataset available on Kaggle
This work is inspired by the potential of GANs in medical image synthesis and contributes to the ongoing efforts in breast cancer research.