G80.3122-001 -- Fall, 2009

Representation and Analysis of Visual Images

Instructor: Eero Simoncelli
Time: Tue 10-12 am
Location: Meyer Hall, rm 1024
[Take elevator to 10th floor. Phone the "Laboratory for Computational Vision" for access]
Prerequisites: linear algebra, linear systems theory, basic probability/statistics. Some matlab programming experience.

Brief Description: A graduate-level lecture course on theory and tools for representing, manipulating and analyzing visual images. Topics to include: imaging and optics, estimation and representation of position, alignment and displacement estimation, estimation and representation of local orientation, multi-scale image decompositions, statistical image modeling and its use in compression, estimation, enhancement and synthesis. Grades are based on homework, which relies on matlab programming.

Date Topic Handouts/Reading Homework
8 Sep
10-12am		 Intro / the visual (plenoptic) world / imaging
[Lecture slides] 		Course description, Background poll,
Plenoptic chapter
  
15 Sep
10-12pm		 Color representation, Multi-D linear systems
[Lecture slides]		    
22 Sep
10-12pm		 Representation of spatial position, sampling, aliasing,    
29 Sep
10-12pm		 Differentiation of sampled images Short article on discrete differentiation  
6 Oct
10-12pm		 Orientation estimation, orientation tensors
[Lecture slides (last 3 sessions)] 		  hw1, due 10/16
shift.m, cconv2.m
rconv2.m, localOri.m
13 Oct
10-12pm		 Orientation estimation, steerability    
20 Oct
10-12pm		 Matching: Fundamentals for displacement estimation or visual search    
27 Oct
10-12pm		 Coarse-to-fine differential matching
[Lecture slides (last 2 sessions)] 		Book chapter  
3 Nov
10-12pm		 Multi-scale representation: prediction across scale    
10 Nov
10-12pm		 Multi-scale representation- pyramids    
17 Nov
10-12pm		 Multi-scale representation- Wavelets
[Lecture slides (last 3 sessions)] 		   
24 Nov
10-12pm		 Image Statistics - intro   hw2, due 12/8
matchedFilter.mat, imSplit.mat
1 Dec
10-12		 Image Statistics - spectral model    
10 Dec
9:30-11:45pm
Thursday!		 Image Statistics - inference Book Chapter hw3, due 12/18
einstein.pgm
15 Dec
10-12		 Image Statistics - adaptive models
[Lecture slides (last 4 sessions)] 		   

Videos, along with Umesh's notes, are available for most lectures here (local access only - check with me for outside CNS/Psych).

Auxilliary Reading: Image Processing, Computer Vision, Biological Vision

  • Two-dimensional Imaging. Ronald N. Bracewell. Prentice Hall, 1995.
  • Pattern Classification, Duda, Hart and Storck. Wiley, 2001.
  • Computer vision: A modern approach. David Forsyth and Jean Ponce. Prentice Hall, 2003.
  • Digital Image Processing . Bernd Jähne. Springer, 1997.
  • A Wavelet Tour of Signal Processing. Stephane Mallat. Academic Press, 1998.
  • Foundations of Vision. Brian Wandell. Sinauer, 1995.
  • Natural Image Statistics -- A probabilistic approach to early computational vision . Aapo Hyvärinen, Jarmo Hurri, and Patrik O. Hoyer. Springer-Verlag, 2009.

    Background Material (mathematics/engineering)

  • My linear algebra notes: linearAlgebra, least squares estimation.
  • Linear algebra appendix from the PDP series, by Michael Jordan: linearAlgebra
  • My convolution/Fourier notes: linSys
  • Linear Algebra and Its Applications. Gilbert Strang. Academic Press, 1980.
  • Discrete-Time Signal Processing. Alan Oppenheim and Ronald Schafer. Prentice-Hall, 1989.
  • Elements of Information Theory. Thomas Cover and Joy Thomas. Wiley Series in Telecommunication, 1991.
  • Probability, Random Variables, and Stochastic Processes. A. Papoulis. 3rd edition, McGraw-Hill, 1991.
  • Probability and Statistics, Morris DeGroot and Mark Schervish. Addison-Wesley, 2002.