CAP6412 - Advanced Computer Vision

Spring 2005
MW 16:30 - 17:45
CDB 221

Instructor

Dr. Alper Yilmaz
Email: yilmaz at cs.ucf.edu (replace at with @)
Office hours: Mon, 3:00-4:30pm, Wed, 3:00-4:00pm, (CSB 250) 
Phone: 

Course Goal

The course is advanced level. The students are required to know basic concepts in Computer Vision, such as: 

• Imaging Geometry

• Camera Modeling and Calibration

• Filtering and Enhancing Images

• Region Segmentation

• Color and Texture

• Line and Curve Detection

• Shape Analysis

• Stereopsis

• Motion and Optical Flow

• Structure from X

Grading Policy

Assignments: 10%
Programming Projects: 30%
Discussion and Attendance: 20%
Presentations: 10%
Reports 30%

Prerequisites

A good background in calculus, geometry, linear algebra, programming in MATLAB or C, and computer vision.

The University Golden Rules will be observed in this class. Copying or Plagiarism is violation of the Golden Rules.

Reference Text:

There is no specific reference.

Lectures

Lecture 1 (January 10, 2005)

• Introduction, Course Overview and List of papers

Lecture 2 (January 12, 2005) Image Regularization

• Presented by Alper Yilmaz

• D. Tschumperle, R. Deriche, "Vector-Valued Image Regularization with PDE's: A Common Framework for Different Applications ", CVPR 2003.

• Link to presentation

• Homework due 19 January 2005, see presentation

Lecture 3 (January 19, 2005) Video Alignment

• Presented by Yinghua Hu

• Y. Caspi and M. Irani, "Alignment of Non-Overlapping Sequences ", ICCV 2001.

• Link to presentation

• Related links: cap5415 lecture 2 cap5415 lecture 15

• Link to Journal version

Lecture 4-5 (January 24-26, 2005) Object Detection

• Presented by Fahd Rafi

• J. Sivic, F. Schaffalitzky and A. Zisserman, "Object Level Grouping of Video Shots ", ECCV 2004.

• Link to presentation

Lecture 6 (January 31, 2005) Wireless Sensor Networks CHANGED

• Presented by Arslan Basharat

• Peshala V. Pahalawatta, Thrasyvoulos N. Pappas, and Aggelos K. Katsaggelos, "Optimal sensor selection for video-based target tracking in a wireless sensor network ", ICIP 2004.

• Related links: 2003 conference version, Unscented Kalman filters, Kalman filters

• Link to presentation

• Homework due 7 February 2005, Derive eq. (1) in the paper

Lecture 7 (February 2, 2005) Tracking

• Presented by Saad Ali

• L. Torresani and C. Bregler, "Space time tracking ", ECCV 2002.

• Link to presentation

Lecture 8-9 (February 7-9, 2005) Tracking

• Presented by Saad Masood Khan

• D. Ramanan and D. Forsyth, "Finding and tracking people from the bottom up ", CVPR 2003.

• Link to presentation

Lecture 10-11 (February 14-16, 2005) Object Detection

• Presented by Alex Aved

• P. Viola, M. Jones and D. Snow, "Detecting Pedestrians using Patterns of Motion and Appearance", ICCV 2003.

• Link to presentation

• Related Links: Presentation on AdaBoost

Lecture 12 (February 21, 2005) Video Completion

• Presented by Mumtaz Sheikh

• Y. Wexler, E. Shechtman, and M. Irani, "Space-Time Video Completion", CVPR 2004.

• Link to presentation

Lecture 13-14-15 (February 23-28, March 2 2005) Stereopsis

• Presented by Xiaochun Cao

• Steven Seitz, "The Space of All Stereo Images", ICCV 2001.

• Link to presentation

Lecture 16 (March 21, 2005) Object Tracking

• Presented by Jeremy Weinsier

• Lorenzo Torresani, Danny Yang, Gene Alexander, Christoph Bregler, " Tracking and Modelling Non-Rigid Objects with Rank Constraints", CVPR 2001.

• Link to presentation

Lecture 17 (March 23, 2005) Rectification

• Presented by Yinghua Hu

• R. Hartley, " Theory and Practice of Projective Rectification", IJCV, Vol 35, No. 2, 1999.

• Link to presentation

Lecture 18-19 (March 28,30 2005) Epipolar Geometry

• Presented by Alper Yilmaz

Lecture 20 (April 4, 2005) Point Matching

• Presented by Fahd Rafi

• H. Chui and A. Rangrajan, "A new algorithm for non-rigid point matching ", CVPR 2000.

Lecture 21 (April 6, 2005) SFM

• Presented by Saad Masood Khan

• J. Xiao and T. Kanade, "Non-rigid shape and motion recovery: degenerate deformations", CVPR 2004.

• Link to presentation

Lecture 22 (April 11, 2005) Shape recovery

• Presented by Saad Ali

• Kolmogorov and Zabih, "Multi-Camera Scene Reconstruction via Graph Cuts ", ECCV 2002.

Lecture 23 (April 13, 2005) View geometry

• Presented by Arslan Basarat

• M.Irani, T.Hassner, and P.Anandan, "What Does the Scene Look Like from a Scene Point? ", ECCV 2002.

• Link to presentation

Lecture 24 (April 18, 2005) SFM

• Presented by Alex Aved

• E. Shechtman, Y. Caspi and M. Irani, "Increasing space time resolution in video", ECCV 2002.
• Journal version.

Programming Assignments

• Assignment 1:
  Implementation of "Alignment of Non-Overlapping Sequences"
  Deliverables: Detailed report including results of every step of the algorithm (Compensated motions etc.), program source code, project live demo in the class.

• Assignment 2:
  Implementation of either "Tracking and Modelling Non-Rigid Objects with Rank Constraints" or "Space Time Video Completion"
  Deliverables: Detailed report including results of every step of the algorithm, program source code, project live demo in the class.

Source Code:

Presentation and source code on interest point detectors (Harris, SIFT, C Schmid)

cimage.cpp cimage.h

Leading Journals and Conferences in Computer Vision

• International Journal of Computer Vision (IJCV)

• Pattern Analysis and Machine Intelligence (PAMI)

• IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR)

• International Conference on Computer Vision (ICCV)

Some other Links

• UCF Computer Vision Lab Home Page

• Computer Vision Home Page