A learning approach for single-frame face super-resolution | IEEE Conference Publication | IEEE Xplore