Not all pixels are equal: Difficulty-aware semantic segmentation via deep layer cascade

Publication
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

Related