CSE/EE586 Computer Vision II
Mathematical Tools for Computer Vision
CSE Department, Penn State University
Instructor: Robert Collins

Fall 2007 Schedule: Tues/Thurs 8:00-9:15AM EES 116

Course Introduction
Collins Intro to Course (Aug 28)   [slides]   [6 per page]
How to do XYZ (Aug 30)   [slides]   [6 per page]
Paper Search/Reading/Writing Resources
Gaussian Mixture Models and Expectation Maximization
Collins Lecture Notes (Aug 28)
GMM, EM and K-means (Sep 04)   [slides]   [6 per page]
Sample paper presentation. Jones and Rehg, Statistical Color Models with Application to Skin Detection. IJCV 46(1):81-96, Jan 2002.   [slides]   [6 per page]
Sample Critique on Jones and Rehg paper. I would have graded this with a check (on our check-minus, check, check-plus scale)
Estimating Gaussian Mixture Densities with EM - A Tutorial, Carlo Tomasi.
Old and New Matrix Algebra Useful for Statistics, Tom Minka.
Research Papers / Oral Presentations
Tuesday, Sep 11
Gary Mixture Models for Optical Flow Computation, Jepson and Black
Guarav An EM-like Algorithm for Color-histogram-based Object Tracking, Zivkovic and Krose
Thursday, Sep 12
Joao Understanding Background Mixture Models for Foreground Segmentation, Power and Shoonees
Arhan Blobworld: Image Segmentation using EM ..., Carson et.al.
Tuesday, Sep 18
Denise Tracking Colour Objects using Adaptive Mixture Models, McKenna et.al.
EM Project Ideas
- From Bishop book: use kmeans or EM to label/compress colors in an image. Try diff values of K. maybe compare k-means and EM results. Bishop's text is on pages: [428] and [429]
- From Forsyth and Ponce: a number of ideas, including color/texture segmentation (e.g. blobworld), fitting line segments (maybe compare EM with RANSAC?), motion segmentation, background subtraction (see Power and Shoonees paper). See pages: [359], [360], [361], [362], [363], [364], [365], [366], [367], [368].
- Two other project ideas from the last time I taught this course. One involves motion segmentation (I provide a dataset), one involves experimenting with the Jones and Rehg skin color classifier. [here are the ideas].
- Of course, I encourage you to come up with your own project ideas involving research or data that is relevant to you.
Dynamic Programming
Collins Dynamic Programming (Sep 20, 25)   [slides]   [6 per page]
Research Papers / Oral Presentations
Thursday, Sep 27
Duane Image Quilting for Texture Synthesis and Transfer, Efros and Freeman
James Seam Carving for Content-Aware Image Resizing, Avidan and Shamir
Tuesday, Oct 02
Jesse Intelligent Scissors for Image Composition.pdf, Mortensen and Barrett
 
Procrustes Shape Analysis
Collins Procrustes Analysis (Oct 4)   [slides]   [6 per page]. Also, some handwritten notes on centers of mass alignment are scanned here .
Collins Review of PCA (Oct 9)   [slides]   [6 per page]
Collins Active Shape Models (Oct 11)   [slides]   [6 per page]
- Least-Squares solution for rotation, translation and scale that aligns two point sets is derived in Efficient Calculation of Absolute Orientation with Outlier Rejection, Bandari etal. Note, this solves for 3D point sets, but same idea should be applicable in 2D (and maybe even simplifies somewhat). See also the alternative approach in Tim Cootes' papers.
- Fun Reading: The Shape of Madness, Mackenzie, Discover Magazine
Research Papers / Oral Presentations
Tuesday, Oct 16
I-Hsien Color active shape models for tracking non-rigid objects, Koschan etal
Ankur Quantifying biomechanical motion using Procrustes motion analysis, Adams and Cerney
Thursday, Oct 18
Sana Training models of shape from sets of examples, Cootes etal. A related paper on fitting them to images is Active Shape Models - Smart Snakes
Brett Shape and the information in medical images: a decade of the morphometric synthesis, Bookstein
 
PCA/LDA/ICA
Collins PCA (Oct 23)   [slides]   [6 per page]
Scribbled notes on PCA vs LDA and ICA. Sorry they are so messy!
Collins PCA/LDA/ICA for Face Recognition (Oct 25)   [slides]   [6 per page]
 
Graphical Models
Collins Intro to Graphical Models (Oct 30)   [slides]   [6 per page]
Reading: Chapter 8, Graphical Models, PRML book, Chris Bishop.
Collins Message Passing (Nov 6)   [slides]   [6 per page]
sumProductChain.m sample code for sum-product algorithm on chains
maxProductChain.m sample code for max-product algorithm on chains (updated Nov8)
Factor graph example 8.51 From Bishop's book   [slides]   [6 per page]
factorgraph.m code for the factor graph example in Bishop's book
Collins Hidden Markov Models (Nov 8)   [slides]   [6 per page]
HMM Reading: A Tutorial on Hidden Markov Models, Rabiner
HMM Research Papers / Oral Presentations
Tuesday, Nov 13
Sana Bayesian Computer vision System for Modeling Interactions, Oliver et.al.
Jonathan Hidden Markov Models for Face Recognition, Nefian and Hayes
Denise Face Segmentation for Identification using Hidden Markov Models, Samaria
Thurs, Nov 15
Gary Real-time American Sign Language Recognition, Starner et.al.
Greg Video-based Face Recognition using Adaptive Hidden Markov Models, Liu and Chen
 
Graphical Models II
Hundal Guest lecture, Intro to Entropy and Gibbs Distributions (Nov 27)   [scanned notes]
Hundal Guest lecture, Gibbs Distributions and GRFs (Nov 29)   [scanned notes]
Collins Intro to Markov Random Fields (MRF) (Dec 4)   [scanned notes]
-movie- Huttenlocher lecture on Efficient BP (Dec 6)   [slides]
MRF Research Papers / Oral Presentations
Tuesday, Dec 11
Joao Interactive Graph Cuts fpr Optimal Boundary and Region Segmentation, Boykov and Jolly.
Mohit An Application of Markov Random Fields to Range Sensing, Diebel and Thrun.
Guarav Comparison of Graph Cuts with Belief Propagation for Stereo, Tappen and Freeman.
Thursday, Dec 13
James Learning 3D Scene Structure from a Single Still Image, Saxena, Sun and Ng.
Ankur Consistent Segmentation for Optical Flow Estimation, Zitnick, Jojic and Kang.