Unsupervised Object Segmentation in Video by Efficient Selection of Highly Probable Positive Features.

Emanuela Haller, Marius Leordeanu

Browse the full ICCV paper archive.