15-385/685 Computer Vision

Carnegie Mellon University

Spring 2011 School of Computer Science

Course Description

An introduction to the theory and practice of computer vision, i.e. the analysis of the patterns in visual images with the view to understanding the objects and processes in the world that generate them. Major topics include optics, image representation, feature extraction, image processing, object recognition, feature selection, probabilistic inference, perceptual analysis and organization, dynamic and hierarchical processing. The emphasis is on the learning of mathematical concepts and techniques and the translation of them to Matlab programs to solve real vision problem. The discussion will be guided by comparision with human and animal vision, from psychological and biological perspectives. Prerequisites: 15-100, 21-120 or permission of instructor. 21-241 preferred but not required.

Course Information

Instructors	Office (Office hours)	Email (Phone)
Tai Sing Lee (Professor)	Mellon Inst. Rm 115	tai@cnbc.cmu.edu
	GHC 8017 (Tu/Thur 4:30-5:30 pm), MI 115 (Wed 9:30-10:30)
Mabaran Rajaraman (385 TA)	W 3:30-4:30 (every week), F 4:30-5:30 (every other week) B 127 Hamerschlag	mabaran@cmu.edu
Ben Poole (385 TA)	M 4:30-5:30 (every week). F: 4:30-5:30 (every other week) GHC 5201 or 5205 clusters	ben@cmu.edu

* Time and Place of Office Hours subject to change. Will announce if change.

Class location and time: Wean 7500. Tuesday, Thursday 10:30 a.m - 11:50 a.m.
Website: http://www.cnbc.cmu.edu/~tai/cv11.html (course info)
Course directory: /afs/andrew.cmu.edu/scs/cs/15-385 (some assignments materials)
Blackboard: http://www.cmu.edu/blackboard/ (Both 385/685 students should use 385 BB for access of course materials and announcements. 685 students should also check their 685 BB for additional requirements

Classroom Etiquette

Please turn OFF your laptop, cell phones or any other electronic devices in the classroom or you will be asked to leave the classroom. Thank you.

Grading Scheme

Evaluation	% of Grade
4 Assignments	50
Quiz 1	10
Term Project	25
Quiz 2	15
685	see additional requirement

Homework Assignments

You will have about 2 weeks to do each homework assignment. Homework report should be type-written. You should submit a hardcopy, documenting all the answers and results in class on the due day, as well as a softcopy of write-up and the codes to Blackboard before 10:00 a.m. on the due day. Blackboard submission after 10:00 a.m. is considered late.
385 students are allowed to collaborate in the two first assignments and in the term project, but not in the third and fourth assignments. Each should submit a copy of the solution, but should list his/her partner, if any, on the first page of the report.
685 students need to complete all the work required of 385, and compete at the same scale. They have to complete the additional challenge questions in each homework (no collaboration). The total grade for these additional questions for all the assignments have to be over 80 percent to be considered satisfactory.

Late Policy

Each student has ONE chance to turn in one of four the homework assignments late within one week without penalty. However, Blackboard submission after noon on the due day is considered late.
If you need to submit a late homework after using the one free late, you will receive 5 percent deduction of your maximum credit per lecture until the solution set is released.
No extension possible for the term project.

Term project

385 students have to do a term project. They might work in pair.
685 students have to do a term project individually, no partner allowed. In addition, their projects need an extensive background review.
The term project will consist of a (1 page) proposal, a (1 page) progress report, and a (up to 6 page) final report, in the CVPR format provided. Matlab codes and additional output should also be submitted as supplementary materials in a different pdf/doc file and/or matlab files.
The projects will be presented on the day of the Final Exam. Format to be defined.

Examinations

There will be two quizzes to test understanding of the materials covered and discussed in the lectures.

Additional Requirement for 685 students

(1) 685 students have to complete the additional challenge questions in each homework (no collaboration). The total grade for these additional questions for all the assignments have to be over 80 percent to be considered satisfactory. In the event of unsatisfactory performance, 685 students will be given a second chance to make up one challenge homework by doing an instructor specified challenge during the term project period.
(2) Their project is expected to be more substantial, and the term paper/report is expected to be accompanied by a in-depth literature review like a conference paper.

Final Grade Assignment

385 students: A for total grade >= 88 %, B: < 88 AND >= 75, C: < 75, >=60. F for cheating.
685 students: A for total grade >= 88 % AND satisfactory performance of the additional assignments.

Laboratory/Help Sessions

Matlab tutorial sessions:
Hunt Lower Level Near/Far 5:30 PM-7:00 PM 01/20/2011
Hunt Lower Level Near/Far 5:30 PM-7:00 PM 01/26/2011
Lab/Help sessions: The teaching assistants will hold "Lab/Help" sessions every other week, at the same place, and at the same time, to help students with programming issues as well as homework assignments.

Syllabus

* Relevant Reading: Richard Szeliski, Computer Vision: Algorithms and Applications . Springer, 2010.

Date	Lecture Topic	Relevant Readings	Assignments/Help Sessions
	IMAGE AND FEATURE REPRESENTATION
T 1/11	1. Introduction to computer vision	ch. 1
R 1/13	2. Optics and Image formation	ch 2.2
T 1/18	3. Depth perception	ch 2.1, 11.3,11.4
R 1/20	4. Stereo computation	ch 2.1, 11.3, 11.4	HW 1 out. Matlab basics 5:30-7:00
T 1/25	5. Linear Filters	ch 3.2, 3.4
R 1/27	6. Laplacian pyramid	ch 3.5, handouts
T 2/1	7. Fourier and Wavelet Analysis	ch 3.3	HW 1 due. HW 2 out;
R 2/3	8. Edge and line detection	ch 4.2, 4.3, 5.1.1
	OBJECT AND SCENE RECOGNITION
T 2/8	9. Principal component analysis	Appendix A
R 2/10	10. Feature extraction (ICA and LDA)	handouts	HW 2 due.
T 2/15	11. Bayesian Classifcation	Appendix B	HW 3 out
R 2/17	12. Wavelets and texture	ch 3.5, 10.5
T 2/22	13. Object detection (boosting)	ch14.1
R 2/24	14. Visual words and recognition	ch 4.4	Preliminary Project Proposal
T 3/1	15. Active deformable models	ch 14.2
R 3/3	Quiz 1 (up to Lecture 13)		Project ideas
M 3/7	Midterm Grade due 6 p.m.
T 3/8	Spring break
R 3/10	Spring break
T 3/15	16. Motion and optical flow	ch 8.4	HW 3 in, HW 4 out ;
R 3/17	17. Object tracking	ch 8.5
	PERCEPTUAL INFERENCE
T 3/22	18. Lightness and color	ch 2, 3.1
R 3/24	19. Surface interpolation (MRF)	ch 3.7	HW 4 due
T 3/29	20. Segmentaton methods	ch 5.4-5.5
R 3/31	21. Shape from shading	ch 12
T 4/5	22. Shape from images	ch12, handouts
R 4/7	23. Structure from motion	ch 8.4, 8.5, ch 7
T 4/12	24. Context and scene recognition	ch 14.4	Project midterm (3)
R 4/14	Spring Carnival
T 4/19	25. Dynamic inference	handouts
R 4/21	26. Hierarchical/Interactive systems	handouts
T 4/26	Project Presentation
R 4/28	Project Presentation		All term papers and projects due (20)
M 5/9 WEH 7500	Final Exam Day: Quiz 2 & Project Presentation
R 5/12	Final Grade due 6 p.m.

Homework Assignments

Homework Assignment 1: Stereo computation
Homework Assignment 2: Pyramid and Image Mosaic
Homework Assignment 3: Texture classification
Homework Assignment 4: Object Tracking

Questions or comments: contact Tai Sing Lee
Last modified: May 15, 2011, Tai Sing Lee

15-385/685 Computer Vision

Carnegie Mellon University

Spring 2011 School of Computer Science

Course Description

Course Information

Recommended Textbook

Classroom Etiquette

Grading Scheme

Homework Assignments

Late Policy

Term project

Examinations

Additional Requirement for 685 students

Final Grade Assignment

Laboratory/Help Sessions

Syllabus

Homework Assignments