X-Ray Medical Image Transformation using StarGAN v2

This repository contains the implementation of X-Ray medical image transformation using StarGAN v2. StarGAN v2 is employed for its advanced capabilities in image translation tasks.

Disclaimer

This repository is for educational purposes only. The utilization of the StarGAN v2 model is solely for academic exploration and understanding of image transformation techniques.

Usage

1. Resize Images: Use the following command to resize images before training:

   python resize.py

2. Train the Model: Train the model with the following command:

   python main.py --mode train --num_domains 3 --w_hpf 0 --lambda_reg 1 --lambda_sty 1 --lambda_ds 2 --lambda_cyc 1 --train_img_dir data/x-raymed/train --val_img_dir data/x-raymed/val --checkpoint_dir expr/checkpoints/x-raymed --sample_every 1000 --save_every 5000

3. Generate Reference Images: Generate reference images using the trained model:

   python main.py --mode sample --num_domains 3 --resume_iter 5000 --w_hpf 0 --checkpoint_dir expr/checkpoints/x-raymed --result_dir expr/results/x-raymed --src_dir assets/representative/x-raymed/src --ref_dir assets/representative/x-raymed/ref

4. Evaluate and Generate Single Images: Evaluate and generate single images using the trained model:

   python main.py --mode eval --num_domains 3 --w_hpf 0 --resume_iter 5000 --train_img_dir data/x-raymed/train --val_img_dir data/x-raymed/val --checkpoint_dir expr/checkpoints/x-raymed --eval_dir expr/eval/x-raymed 

For more detailed instructions and options, please refer to the documentation.