CS294-6: Theoretical, Conceptual and Experimental Vision

Time: Mon & Wed, 13:10-14:30. Fall, 2006.
Location: 405 Soda Hall.
Units: 3.
Instructors: Ruzena Bajcsy, Shankar Sastry, and Allen Yang.
Emails: { bajcsy, sastry, yang } @ eecs
Office hour:     Mon from 2:30-3:30 after class or by appointment for Prof. Bajcsy. 665 Soda Hall.
                      Tue & Thu from 3:00-4:00 or by appointment for Prof. Sastry. 284 HMMB.
                        Tue from 2:30-4:00 for Allen. 343 Cory Hall (TRUST center).

Goal:

The objective of this course is to give an introduction to some basic concepts and foundations of computer vision with emphasis on geometric methods in vision such as Structure from Motion and Generalized Principal Component Analysis (GPCA). We will discuss both concepts, theory and implementation and experimental aspects. An important aspect of the course will be the development of active vision, which consists of closing the loop around the camera sensor. Scenarios that we will cover include mobile robots (both driving and flying), manipulating objects and vision-based communication between people and machines.

The aim of the course is to enable students to set up experimental stereo camera systems, evaluate their characteristics, and use the latest geometric methods to calibrate them and reconstruct the motion of single and multiple objects.

Required Textbook:

An Invitation to 3D Vision. By Yi Ma, Stefano Soatto, Jana Kosecka , S. Shankar Sastry. Springer Verlag, 2005.

Website for the book : http://vision.ucla.edu/MASKS/. Website for the sample codes: http://cs.gmu.edu/~kosecka/bookcode.html.

Grading Policy:

There will be approximately 5 problem sets (50% credit) and a final project (50% credit) for the class, Participation in the class is expected from the registered students. Homework is due at the beginning of the class on the due date.

Course Projects Final Reports: (Thank you all for your slides!)

Course Progress:

Date	Topics	Downloads	HW
Week 1	Introduction and image formation	outline,lecture1.
Week 2	Optics, radiometry and error analysis	Optic and radiometry.lecture2; Sensor errors.lecture3.	HW 1 issued
Week 3	Image primitives and correspondence	lecture 4
Week 4	Review of basic algebra and geometry	lecture 5, lecture 6	HW1 due Sep 20. HW 2 issued.
Week 5	Two-view geometry	lecture 7, lecture 8, lecture 9.	HW2 due Sep 27. HW3 issued.
Week 6	Camera calibration	lecture 10, lecture 11, lecture 12
Week 7	Multiple-view geometry Structure from symmetry	lecture 13 & 14 lecture 16	HW 3 due Oct 11 HW 4 issued. Matlab files: 1, 2.
Week 8	Active vision	lecture 15.	Mid-proposal due Oct 18
Week 9	Actuve vision	lecture 17.
Week 10	Visual feedback	lecture 18, lecture 19.	HW 4 due Nov 1
Week 11	Real-time vision	lecture 20. lecture 21	HW 5 issued.
Week 12	Brief project overview. Step-by-step reconstruction (by Prof. Soatto)
Week 13	GPCA: Introduction Visual Compass (by Prof. Kosecka)	lecture 23.	HW 5 due Nov 22
Week 14	GPCA: Algebra	lecture 25
Week 15	GPCA: Robust statistics	lecture 26

Additional Material:

Other textbooks closely related: Multiple View Geometry in Computer Vision by Richard Hartley and Andrew Zisserman.
Reading on affine local feature detection and matching: [Mikolajczyk, Lowe, Nister(pdf, ppt)].
Reading on sensor errors: [Kamberova].
Reading on feature tracking: [Shi, Tommasini].
Reading on approximate camera models and reconstruction: Orthographic [Tomasi]; Paraperspective [Aloimonos, Poelman]; and Affine [Kanatani].
Reading on robust vision techniques: [Torr, Steward].
Reading on active vision: [Bajcsy, Krotkov, Krotkov, Madden, Anderson, Uhlin].
A textbook draft on GPCA: see Prof. Ma's course website.
Reading on robust GPCA: [Yang, Yang]. See also http://perception.csl.uiuc.edu/gpca/