grid2vec: Learning Efficient Visual Representations via Flexible Grid-Graphs

Publication
CoRR