TR-IIS-06-008    Fulltext

Analysis of Sampling-based Texture Synthesis as a Generalized EM Algorithm

Liu-yuan Lai, Wen-Liang Hwang, Silong Peng


Research on texture synthesis has made substantial progress in recent years, and many patch-based sampling algorithms now produce quality results in an acceptable computation time. However, when such algorithms are applied to textures, whether they provide good results for specific textures, and why they do so, are questions that have yet to be fully resolved. In this paper, we deal specifically with the second question by modeling the synthesis problem as learning from incomplete data. We propose an algorithm and show that the solution of many sampling-based algorithms is an approximation of finding the maximum-likelihood optimum by the generalized expectation and maximization (EM) algorithm.

Keywords: Texture synthesis, EM algorithm