This is the first introduction course in computer vision offered to undergraduate students. Most students in the class are juniors and seniors but sophomores and graduate students are also found in the class.
However, I personally found that CS 2800 was not used too often. Instead, some background in Linear Algebra (MATH 2940, 2210, etc.) would be helpful as much of the concepts are founded on matrix operations.
- Image filtering
- Edge detection
- Image Resampling
- Image interpolation
- Feature Detection
- Image Transformations
- Facial and Object Detection
Expect 4-5 projects for the semester, one mid-term (take home during the FA2012 semester), and one final. Projects are milder than the likes of CS 3110 - Data Structures and Functional Programming. The projects do not involve as much coding as other project-based courses. Instead, the projects tend to focus on the math and vision algorithms covered in class. Expect poor documentation of the many libraries used in the projects.
There are extra credits in this class - if you have time, try to do some of them. You’ll learn a lot more since you’ll either a) get to implement something you haven’t been taught or b) get to implement something you normally wouldn’t have time for.
CS 4670 is very similar to the CV course at the University of Washington, taught by Steve Seitz. Some projects use the same sample images and many of the presentations share similar slides. You won’t find already implemented projects/solutions but it is worth noting the class curriculum is very similar to the UW class.
|Semester||Time||Professor||Median Grade||Course Page|
|Fall 2012||MWF 10:10 - 11:00||Noah Snavely||B||http://www.cs.cornell.edu/courses/cs4670/2012fa/|
|Fall 2010||MWF 10:10 - 11:00||Noah Snavely||A-||http://www.cs.cornell.edu/courses/cs4670/2010fa/|