Learning Mixtures of Bernoulli Templates by Two-Round EM with Performance Guarantee
This provides a theoretical guarantee for learning mixtures of Bernoulli templates, which model images in computer vision, but it is an incremental extension of prior Gaussian mixture work.
The paper generalizes a two-round EM algorithm to learn mixtures of high-dimensional Bernoulli templates, showing it achieves near-optimal precision with high probability under conditions of sufficient template difference and feature count.
Dasgupta and Shulman showed that a two-round variant of the EM algorithm can learn mixture of Gaussian distributions with near optimal precision with high probability if the Gaussian distributions are well separated and if the dimension is sufficiently high. In this paper, we generalize their theory to learning mixture of high-dimensional Bernoulli templates. Each template is a binary vector, and a template generates examples by randomly switching its binary components independently with a certain probability. In computer vision applications, a binary vector is a feature map of an image, where each binary component indicates whether a local feature or structure is present or absent within a certain cell of the image domain. A Bernoulli template can be considered as a statistical model for images of objects (or parts of objects) from the same category. We show that the two-round EM algorithm can learn mixture of Bernoulli templates with near optimal precision with high probability, if the Bernoulli templates are sufficiently different and if the number of features is sufficiently high. We illustrate the theoretical results by synthetic and real examples.