Outline Fall 2013

CS9840a Learning and Computer Vision

Course Description

Traditionally, researchers in the field of computer vision have been hand-crafting appropriate physical/statistical models of objects/natural scenes for building computer vision systems. Recent advances in imaging and computing technology make it possible to capture and process large amounts of visual data efficiently. This lead to increasing use of machine learning techniques for model learning in computer vision. A model learned from large visual datasets is less likely to be brittle than a model hand-crafted by a designer. In this course, we will explore recent successful computer vision methods based on machine learning. The course will be organized as a combination of lectures by the instructor and paper presentation by the students. After we study a new machine learning technique, we will read and discuss a paper that makes use of that technique. Each student will do a paper presentation, as well as a programming project.


Machine Learning topics will be selected from the following:

  • Nearest Neighbor Classifiers
  • Linear Classifiers
  • Neural networks
  • Ada Boost
  • Dimensionality reduction with PCA
  • Expectation Minimization (EM)

Computer Vision applications will be selected from the following:

  • Image Segmentation
  • Motion segmentation
  • Object Recognition
  • Action Recognition

Prerequisites

A course on computer vision or image processing; strong programming skills in C or C++; familiarity with statistics, calculus, linear algebra. Students lacking these requirements should speak with the instructor for obtaining permission to enroll.

Instructor

Olga Veksler
Office: MC361
Email: my_first_name@csd.uwo.ca
Lectures: Wednesday 10:30-12:30
Lecture Room: MC320
Office Hours: Wed 2:00-3:00 or by appointment.

Textbook

There will be no required textbook in this course. The course will be based on papers that I will hand out for reading. For reference, students can use the following books:
  • Richard Szeliski, Computer Vision: Algorithms and Applications
  • Forsyth and Ponce, Computer Vision -- A Modern Approach
  • Trucco and Verri, Introductory Techniques for 3-D Computer Vision
  • Duda and Hart and Stork, Pattern Classification
  • Tom Mitchell, Machine Learning

Course Website

http://www.csd.uwo.ca/~olga/Courses//Fall2013//CS9840/index.html

Student Evaluation

Grades will be based on:
  • Class participation, 10 % of the final mark
  • In class paper presentation, 30 % of the final mark
  • Final project presentation, 20 % of the final mark
  • Written report and the code for the final project, 40 % of the final mark

Final Project

A student will chose a final project in consultation with me. The project must be related to computer vision and must make use machine learning techniques. Students should select their final project topic by November 2, and the final project written report and the code is due on January 20.

Ethical Conduct

Plagiarism:Students must write their essays and assignments in their own words. Whenever students take an idea, or a passage from another author, they must acknowledge their debt both by using quotation marks where appropriate and by proper referencing such as footnotes or citations. Plagiarism is a major academic offence (see Scholastic Offence Policy in the Western Academic Calendar).