Date: Tuesday 13 November 2007, 13:55
Venue: EM3.02
Name: Dr. Ioannis Ivrissimtzis, Durham University
In statistical modeling, the model parameters are usually estimated by maximizing a probability measure, such as the likelihood or the posterior. In contrast, variational Bayesian methods treat the parameters of a model as probability distributions and compute optimal distributions for them. We applied a variational Bayesian technique to surface fitting for height field data. A series of uniform bicubic B-spline surfaces with increasing numbers of control points are fitted on a local neighborhood of the data, and the model with the highest evidence, is selected. As an application, the basic surface fitting algorithm is used for point denoising.