Advanced Computer Vision

4003-558-01
4005-758-01

 

TEXTBOOK:
Digital Image Processing USING MATLAB   Gonzalez, Woods and Eddins
Second Edition  - Gatesmark Publishing  (c)2009

MATLAB RESOURCES:
Mathworks provides tutorials online:
http://www.mathworks.com/academia/student_center/tutorials/launchpad.htmlYOU CAN ACCESS MATLAB ON THE CS COMPUTERS  IF YOU HAVE ANY PROBLEMS RUNNING MATLAB CONTACT THE CS ADMIN OFFICE

HOMEWORK POLICIES:

1. Homework assignments can be done individually or teams of two.
2. Assignments are due at 10am. Email material to the course account  rsg.advcv@gmail.com


COURSE GRADING POLICIES - Subject to change:
Course grades will be based on homework assignments, in class discussions and final group project
Note: There are no written exams in this course, but you will be graded on your class participation (discussing homework assignment, assigned papers, etc.). Class attendance is required.
All group projects will be done in Matlab - no exceptions


Homework:                                            25%
Class discussions:                                  25%
Project report and Presentation:             50%

Papers of potential interest:
1. Kernel Codebooks for Scene Categorization       Jan C. van Gemert, Jan-Mark Geusebroek, Cor J. Veenman, and Arnold W.M. Smeulders http://dare.uva.nl/document/126930
indoor_camera_ready.dvi

2. Recognizing Indoor Scenes  Ariadna Quattoni and Antonio Torralba      http://eprints.pascal-network.org/archive/00005818/01/cvpr09.pdf

3. Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories  Svetlana Lazebnik, Cordelia Schmid and Jean Ponce

4. final_version_co_clustering-Shah.dvi Scene Modeling Using Co-Clustering final_version_co_clustering-Shah.dvi Jingen Liu and Mubarak Shah

5.  Spatial hierarchy of textons distributions for scene classificationby    Sebastiano Battiato





Tentative Schedule:

              
Class 01
11/29 Tu
                   
Lecture 1      
Outline Presentation

Homework #1 - DUE THURSDAY,  December 1st  - see lecture notes
Homework #2 - DUE TUESDAY,   December 6th - see lecture notes
Paper: How many pixels make an image? "How many pixels make an image?," Antonio Torralba, Visual Neuroscience (2009), 26, 123-131     How many pixels make an image?
How many pixels make an image? Purple Flower
  
Class 02
12/01 Th
Lecture 2
HW#1 Due
Video Lecture: http://videolectures.net/nips09_torralba_uvs/
Class 03
12/06 Tu
HW#2 Due
Resolution presentations and discusions - Resolution Response Form
Paper: "What makes an image memorable? "  Phillip Isola Jianxiong Xiao Antonio Torralba Aude Oliva
Class 04
12/08 Th
Lecture 3
Flower2A.JPG

Class 05
12/13 Tu
 Lecture 4
HW#3 Due






Class 06
12/15 Th
 10-15 minute  Presentations

Class 07
01/10 Tu
 Yuheng Ph.D. research on video scene categories


Class 08
01/12 Th
 No lecture/class - Teams use class time to formalize project - write project proposals due January 17
Class 09
01/17 Tu
 Team project presentations
Class 10
01/19 Th

Class 11
01/24 Tu

Class 12
01/26 Th





Class 13
01/31 Tu

Class 14
02/02 Th

Class 15
01/07 Tu

Class 16
02/09 Th

Class 17
02/14 Tu

Class 18
02/16 Th

Class 19
02/21 Tu




Class 20
02/23 Th
 Guest Speaker