Multiscale Conditional Random Fields for Semi-supervised Labeling and Classification

David Duvenaud, Benjamin Marlin, Kevin Murphy
2011 2011 Canadian Conference on Computer and Robot Vision  
Motivated by the abundance of images labeled only by their captions, we construct tree-structured multiscale conditional random fields capable of performing semisupervised learning. We show that such caption-only data can in fact increase pixel-level accuracy at test time. In addition, we compare two kinds of tree: the standard one with pairwise potentials, and one based on noisy-or potentials, which better matches the semantics of the recursive partitioning used to create the tree.
doi:10.1109/crv.2011.56 dblp:conf/crv/DuvenaudMM11 fatcat:r6qlr7ee4fg7zoq6fqhxyrtszy