Efficient computation of robust low-rank matrix approximations in the presence of missing data using the L1 norm.

Anders P. Eriksson, Anton van den Hengel

Browse the full CVPR paper archive.