In this paper we introduce a new class of image models, which we call dynamic trees or DTs. A dynamic tree model specifies a prior over a large number of trees, each one of which is a tree-structured belief net (TSBN). Experiments show that DTs are capable of generating images that are less blocky, and the models have better translation invariance properties than a fixed, "balanced" TSBN. We also show that Simulated Annealing is effective at finding trees which have high posterior probability.
|Title of host publication||Advances in Neural Information Processing Systems 11 (NIPS 1998)|
|Number of pages||7|
|Publication status||Published - 1999|