Preview of the cast suppression tool.
The Cycle-GAN model source training code was obtained from the following GitHub repository: https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix
Paper: https://academic.oup.com/jamia/advance-article-abstract/doi/10.1093/jamia/ocab192/6382240
The "Data" folder contains two subdirectories:
- "generated" subdirectory that contains images with the removed cast by the developed Cycle-GAN model.
- "real" subdirectory that contains original images with cast applied. From these images, the images in the "generated" subdirectory are created.
In the archive "Data.zip," both subdirectories are compressed. This archive needs to be uploaded to the "Cast_tool_preview.ipynb" on the Google collab platform if you want to test the tool by yourself.
In the "Notebook" folder, you can find "Cast_tool_preview.ipynb" that is ready to be run on the Google collab platform. You can follow the instructions in the notebook to make it run (For this purpose, we have also provided a ".html" version of the notebook." If you want to have smoother performance, you can run "Cast_tool_previewOpenCV.ipynb" locally on your PC.
Last but not least, you can get a quick preview of the proposed cast removal tool in the video "Cast_tool_preview.mp4," which can be found in the "Video" directory.