Semantic segmentation for medical imaging

What would be the best way to deploy a semantic image segmentation project detecting lung vs heart muscle for medical imaging?

Roboflow doesn’t currently support semantic segmentation but you could frame this as an instance segmentation project as described here: Instance Segmentation Projects

Yes, sry. Let’s say as a general CV Q. What deployment method do you think would make sense for this type of user or project? (Raspberry Pi, OAK, iPad, maybe a different device?)

As with most projects, which device you should choose depends on a lot of factors. Some of the things you might wnat to consider are described here: How to choose your deployment method