Supervised Denoising Challenge AI4Life-MDC25

Example submission container

Welcome to the AI4Life-MDC24 denoising challenge! Check The Challenge Page for all the details about the challenge.

On this page, you can find an example submission for the Grand Challenge platform.

Submission checklist

• Check that all the requirements are contained in the requirements.txt and Dockerfile.
• Test your container locally with do_test_run.sh.
• Create a gzip archive from your image, run do_save.sh. This may take a while!
• Go to this page for further instructions on how to submit to the Grand Challenge.

How to use this code

1. Example inputs are stored in the test/input/ folder.
2. Look through the contents of the inference.py script.
3. Run do_test_run.sh to build and test the container execution.
4. The resulting image should appear in the test/output/ folder.

What is in the example?

Here, we are showcasing an example pytorch model and its inference. The model contains only a Gaussian Blur operator. The model is packaged into jit. See create_model.py for details.

The container runs inference.py script, which loops through the noisy images in the INPUT_PATH and applies the model to them individually. The result denoised images are then saved into OUTPUT_PATH folder.

Useful links

Make sure to check Grand Challenge documentation with any questions you may have.

For any other questions or issues, create a topic on the challenge forum or drop us an email through the Email organizers button on the challenge page.

Thank you for participating, and we are looking forward to receiving your submission!