Should lower and upper models be separately trained?

Since there're upper jaws and also lower jaws of the same patient, so I'm wondering about the training method, should we separately train two models or should we train one model that can process both lower and upper jaws. I think that depends on the testing procedure, if we can get access to the "lower" or "upper" label for the input object, then separate training should be a better choice.

Hi,

Yes, you will get access to the information about the jaw (uppper/lower). You can check the code base (process.py) we pushed to the github of the challenge. https://github.com/abenhamadou/3DTeethSeg22_challenge

Best regards