S2DNet: Learning Image Features for Accurate Sparse-to-Dense Matching.

Hugo Germain, Guillaume Bourmaud, Vincent Lepetit

Browse the full ECCV paper archive.