Computer Vision in a Nutshell

--An 8 Week Reading Group on Selected Topics in Computer Vision

 

Announcement:

1.       The CVinaNutshell is now part of the PhD monitoring.

2.       May/25: I will be out of town next Wednesday. Nianjun will give a lecture on SVM.

 

 

General Information:

Instructors

Yi Li (yi dot li@nicta) and Xuming He (xuming dot he@nicta)

Location

Level 3, Seminar Room D

Time

Wednesday  3pm-4:30pm+/-30 mins

Textbook

Computer Vision: A Modern Approach (CVMA)

Computer Vision: Algorithms and Applications (CVAA)

Multiple View Geometry (MVG)

Pattern Classification (PC)

Pattern Recognition and Machine Learning (PRML)

Acknowledgement

The materials in this syllabus are primarily from David Jacob’s course series and Yiannis Aloimonos’s seminars taught in the University of Maryland.

 

 

Objective:

 Computer vision is a fast growing research area. Each year, hundreds of papers are published in all possible topics in computer vision. The goal of this reading group is to show the big picture of the state of the art of computer vision to the students in NICTA. Due to the limited time, only a number of topics will be covered. Possible topics are:

1.       Camera and Image Formation

2.       Early Vision

3.       Image Segmentation

4.       Face/object Recognition

5.       Image Processing fundamentals

6.       Basic techniques in Computer Vision

 

 

Reading Group Requirements:

 

Homework

We will present one topic in each reading group. For each topic, the students will be required to read the designated chapters and up to 2 selected papers. The students are required to write a one page summary of any of the paper.

There will be a grading on the summary.

Plagiarism

All assignments must be done independently. You are allowed to discuss with your colleagues and the instructors. It is best to understand the problem by your own, because this reading group is solely for your future.

Absences

Once agreed to attend the reading group, you are required to email / talk to the instructors if you are not able to show up.

 

 

Schedule

Date

Instructor

Topic

Required Reading

Background Reading

 

 

 

 

May/18

 

 

 

 

YL

 

 

Introduction

Image Formation

Early vision

Cognition in psychology

Face recognition (I)

 

Lecture Notes

 

Biederman, I. (1987). Recognition--by--components: A theory of human image understanding. Psychological Review, 94(2):115--147.

 

Sinha  et al, ``Face recognition by humans: twenty results all computer vision researchers should know about.’’

 

Turk, M. & Pentland, A. (1991). Eigenfaces for recognition. Journal of Cognitive Neuroscience, 3, 71-86.

 

 

CVMA: Chap. 4, 5, 8

 

CVAA: Chap. 2, 14.1, 14.2

 

PC: Appendix A

 

 

 

 

May/25

 

 

 

 

YL

 

 

 

Image Processing

Diffusion Process

Filter Banks

Noise Removal

 

Lecture Notes

 

 

L. Alvarez, P. Lions, and J.M. Morel.  Image selective smoothing and edge detection by nonlinear diffusion II. SIAM J. of Numerical Analysis

 

Active contour without edges, Chan, T.F.; Vese, L.A., IEEE Transactions on Image Processing, 10 (2), Feb. 2001

 

Blind motion deblurring from a single image using sparse approximation,

J. Cai, H. Ji, C. Liu and Z. Shen, CVPR09.

 

 

CVMA: Chap. 6, 7, 8, 9

 

CVAA: Chap. 3

 

Advanced: Fundamentals of Digital Image Processing, Anil. K. Jain (2 copies available from YL)

 

 

 

 

 

 

Jun/1

 

 

 

 

 

 

 

NJ

 

 

 

 

 

 

Support vector machine

 

 

Martin Szummer, Pushmeet Kohli, Derek Hoiem Learning CRFs using Graph Cuts.In: ECCV 2008.

Catalin Ionescu, Liefeng Bo, Cristian Sminchisescu: Structural SVM for visual localization and continuous state estimation. ICCV 2009: 1157-1164

M. B. Blaschko and C. H. Lampert. Learning to localize objects with structured output regression. In ECCV, pages 2–15, 2008.

The chapters of Statistical Pattern Recognition for SVM by Andrew Webb.
 
I. Tsochantaridis, T. Joachims, T. Hofmann, and Y. Altun, Large Margin 
Methods for Structured and Interdependent Output Variables, JMLR 
 
T. Joachims, T. Finley, Chun-Nam Yu, Cutting-Plane Training of Structural SVMs, Machine Learning 

 

 

Jun/8

 

 

YL

 

Features

Optical flow

Motion

 

 

D. Lowe: Three-Dimensional Object Recognition from Single Two-Dimensional Images.  Artificial Intelligence, 1987.

 

Mikolajczyk and Schmid, A Performance Evaluation of Local Descriptors, PAMI October 2005   pp. 1615-1630

 

 
CVMA: Chap. 4, 6, 8

 

 

 

5

XH

 

Face/Person detection

Face/Object recognition

 

 

Paul Viola, Michael Jones, "Rapid Object Detection using a Boosted Cascade of Simple Features," cvpr, vol. 1, pp.511, 2001

 

Pedro F. Felzenszwalb and Daniel P. Huttenlocher. Pictorial Structures for Object Recognition.  IJCV

 

Stochastic grammar models

 

 

 

6

XH

 

Human pose estimation

Human in street

 

Sidenbladh, H., Black, M. J., and Fleet, D.J., Stochastic tracking of 3D human figures using 2D image motion, ECCV 2002

 

Saul and Roweis: Think Globally, Fit Locally, Unsupervised Learning of Nonlinear Manifolds

 

 

 

 

 

7

XH

 

Scene understanding

Image segmentation

Normalized Cut / Graph cut

Markov random field

J Shi and Jitendra Malik. Normalized cuts and image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence , 22(8):888-905, August 2000.

 

A. Blake, C. Rother, M. Brown, P. Perez, and P. Torr. Interactive image segmentation using an adaptive GMMRF model. Proc. Eur. Conf. on Computer Vision, ECCV (2004)

 

Efficient Graph-Based Image Segmentation

Pedro F. Felzenszwalb and Daniel P. Huttenlocher. International Journal of Computer Vision, Volume 59, Number 2, September 2004

 

Scene understanding papers.