Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in
Computer Vision
Lecture 1: Markov Random Fields
Jens Rittscher and Chuck Stewart
1
Motivation
Up to now we have considered distributions of a single random variable x. Recall
that we wish to be in the position to tre
Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in
Computer Vision
Lecture 6: Markov Chain Monte Carlo and
Gibbs Sampling
Jens Rittscher and Chuck Stewart
Figure 1: Noisy binary image. The left image displays the a binary image containing no noise. The given noisy im
Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in
Computer Vision
Lecture 7: Approximate Inference
Jens Rittscher and Chuck Stewart
1
Overview
Motivation
Iterated Conditional Modes (ICM)
Model Paramters
Pseudolikelihood
2
Motivation
Any property of the posterior
Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in
Computer Vision
Lecture 8: Belief Propagation
Jens Rittscher and Chuck Stewart
1
Overview
In the next several classes we will discuss more recent methods for estimation
using Markov Random Fields. At the same time we
Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in Computer Vision
Lecture 1: Random Variables
Jens Rittscher and Chuck Stewart
1
Motivation
Imaging is a stochastic process:
If we take all the different sources of error into account it is hard to argue
that digital
Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in Computer Vision
Lecture 2: Maximum Likelihood and Bayesian
Estimation
Jens Rittscher and Chuck Stewart
1
Motivation and Problem
In Lecture 1 we briey saw how histograms could be used to model the probability of pixe
Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in Computer Vision
Lecture 4: Gaussian Mixture Models and the EM
Algorithm
Jens Rittscher and Chuck Stewart
1
Motivation
We will continue with our problem of modeling densities.
Similar to kernel density estimates, we
Statistical and Learning Techniques for Computer Vision
CSCI 6966

Fall 2006
Statistical and Learning Techniques in Computer Vision
Lecture 3: NonParametric Density Estimation
Jens Rittscher and Chuck Stewart
1
Overview
Motivation for nonparametric methods
Review of point and histogram techniques
Kernel density methods
Neare