Dear organizers,
I have a couple of questions regarding the model submissions.
From the instructions, I understand that we are required to submit two .sh files, one for CT and one for MRI, for each task. I would like to clarify whether we are expected to submit a single model capable of handling both modalities, or if it is acceptable to submit two separate models, each tailored specifically for CT and MRI, respectively.
Additionally, I have not been able to locate the dataset for Task 3. Could you kindly let me know when it will be made available?
Thanks a lot
Created by Giovanna Nicora gnicora Hi,
To answer your questions, you should use only one model (but might change some hyperparamters for each dataset), and train it on each modality separately, so no need to have one joined model (But please don't use two completely different model/ methods).
For the Task 3, you can use what ever dataset you want to use to train your models to get the best possible performance (as you would to solve a real-world problem), we. dont provide a dedicated one, however you of course can you the brain dataset from task 1&2.
Cheers,
David