Optimizing transformations for contrastive learning in a differentiable framework

Open in new window