Principled fusion of high-level model and low-level cues for motion segmentation.

Arasanathan Thayananthan, Masahiro Iwasaki, Roberto Cipolla

Browse the full CVPR paper archive.