Network from Scientific Research

Use of DLVPM in a Siamese/twin network configuration. Here augmented versions of the same input are fed to a network, and the network is trained to learn DLVs that are invariant to these augmentations.

DOI