Abstract

In this paper, we describe a single frame super-resolution technique for reconstructing high-resolution face images from their very low-resolution counterparts. Our technique is based on learning a subspace for high-resolution images in the transport space.

© 2015 OSA

PDF Article