EE/CNS/CS 148 - Spring 2004

Lectures notes


 





 
 
Lecture 1:
Summary:
Introduction to Recognition - problems to solve, applications. Brief description of the constellation model
03/30/04
Notes
Presentation (pdf,15.9MB)  

 
 
Lecture 2:
Summary:
Matched Filtering
03/30/04
Notes
(pdf,77KB)(ps,110KB)
Matlab Code:
MatchedFilter.m ROC.m

 
 
Lecture 3:
Summary:
Formation of an image - Lambertian surfaces/specular surfaces - RGB images - changes in contrast and brightness and invariance with respect to those - Principal components analysis: singluar values decomposition, computation of the subspace that best represents the data in a given number of dimensions.
04/06/04
Notes
(ps) (pdf)
Matlab Code:
PCAspace.m PCA.m

 
 
Lecture 4:
Summary:
Singular value decomposition and properties - Fisher Linear Discriminants.
04/08/04
Notes:
(ps) (pdf)
Matlab Code:
fisherLD.m fisherDemo.m genData.m
Code and data:
HRL-faces-Belhumeur/

 
 
Lecture 6:
Summary:
The constellation model - case with a single part.
04/15/04
Notes:
(ps) (pdf)

 
 
Lecture 8:
Summary:
Constellation model - case with several parts.
04/22/04
Notes:
(ps) (pdf)
Paper:
ECCV2000 paper by M.Weber (former PhD student in the Vision Lab) - (ps.gz) (pdf)
Matlab code:
lecture8.m gaussian.m randnND.m

 
 
Lecture 9:
Summary:
Constellation model - translation, rotation and scale invariance.
04/27/04
Notes:
(ps) (pdf)
Lecture 15:
Summary:
EM.
05/20/04
Notes:
(Clustering pdf) (EM pdf) (EM Code) (Draw Elipse Code)