Recognizing an image and segmenting it into coherent regions are often
t...
Unsupervised semantic segmentation aims to discover groupings within and...
Weakly supervised segmentation requires assigning a label to every pixel...
The per-pixel cross-entropy loss (CEL) has been widely used in structure...
Existing semantic segmentation methods mostly rely on per-pixel supervis...
We propose a multigrid extension of convolutional neural networks (CNNs)...
We address the problem of converting large-scale high-dimensional image ...