Does it make sense to have a SelfSupervisedSolver class that wraps all cases?
As far as I can see, we could add Noise2Inverse, equivariat imaging, SURE in a loss class that just takes loss(noisy,model) , each with their own parameters to set (e.g. splits, rotations, etc)
Does it make sense to have a SelfSupervisedSolver class that wraps all cases?
As far as I can see, we could add Noise2Inverse, equivariat imaging, SURE in a loss class that just takes
loss(noisy,model), each with their own parameters to set (e.g. splits, rotations, etc)