Multiscale Conditional Random Fields for Vision
By David Duvenaud
Recent work on image classification and segmentation has shown that
incorporating evidence from multiple scales is an effective strategy. The
natural model for sharing evidence at multiple scales is a tree-structured
conditional random field. We develop three refinements to these
tree-structured CRFs: First, we introduce a noisy-or factor that is more
appropriate than standard pairwise factors. Second, we show how to allow
factors to account for variations in graph structure between images. Third,
we show how to do multi-class learning and inference efficiently, as opposed
to one-against-all classifiers.
Visit the LCI Forum page