Unsupervised Semantic Segmentation by Contrasting Object Mask Proposals
Wouter Van Gansbeke1 * Simon Vandenhende1 * Stamatios Georgoulis2 Luc Van Gool1,2
1
KU Leuven/ESAT-PSI 2 ETH Zurich/CVL, TRACE
Abstract
I. Generate Object Mask Proposals Clustering/
Finetuning
Being able to learn dense semantic representations of
images wi ...


雷达卡




京公网安备 11010802022788号







