Computer Vision

Overview

This graduate-level course will explore techniques for the analysis of visual data (primarily color images). In the first part of the course we will examine the physics and geometry of image formation, including the design of cameras and the study of color sensing in the human eye. In each case we will look at the underlying mathematical models for these phenomena. In the second part of the course we will focus on algorithms to extract useful information from images. This includes detection of reliable interest points for applications such as image alignment, stereo and instance recognition; learning representations of images for classification, object detection and denoising; and generative modeling of images. Time permitting we will look at some additional topics at the end of the course.

Course assignments will highlight several computer vision tasks and methods. For each task you will construct a basic system, then improve it through a cycle of error analysis and model redesign. There will also be a final project, which will investigate a single topic or application in greater depth. This course assumes a strong background in basic probability, linear algebra, and ability to program in Matlab or Python. Prior experience in signal or image processing is useful but not required.

Logistics

• Class hours: Tuesday/Thursday 1:00PM - 2:15PM
• Location: Marston Hall, 132
• Moodle for homework and project write-ups.
• Piazza for announcements and discussions.

Grading

• 60% Mini projects (5 in total)
• 15% Weekly homework (10-11 in total)
• 20% Final project
• 5% Class participation

Accommodation statement

The University is committed to providing an equal educational opportunity for all students. If you have a documented physical, psychological, or learning disability on file with Disability Services (DS), you may be eligible for reasonable academic accommodations to help you succeed in this course. If you have a documented disability that requires an accommodation, please notify me within the first two weeks of the semester so that we may make appropriate arrangements.

Academic honesty

Since the integrity of the academic enterprise of any institution of higher education requires honesty in scholarship and research, academic honesty is required of all students. Students are expected to be familiar with this policy and the commonly accepted standards of academic integrity (http://www.umass.edu/honesty).

Acknowlegements

Many of the slides and homework assignments are based on excellent computer vision courses taught elsewhere by Svetlana Lazebnik, Alyosha Efros, Alexander Berg, Steven Seitz, James Hays, Charless Fowlkes, Kirsten Grauman and many others. Many thanks to Richard Szeliski for making the textbook available online for free.