FINO

FINO is a Meta FAIR research framework/method for label-free adaptation of generic vision foundation models to specialized scientific imaging domains using existing metadata as weak guidance together with self-supervised learning.