The class will be mostly project based. Students will be asked to implement a current state-of-the-art vision system and extend the system in novel ways. Students will be broken up into groups of at most 2 for the purpose of the project. Projects will be discussed in class within the first several weeks.
There will be several initial homework assignments designed to give students the necessary background for their project.
Later lectures will consist of a combination of students presenting relevent and current object recognition papers as well as lectures integrating the various approaches presented.
60% Class Project -- Students will work in teams of at most 2 to implement a state-of-the-art visual recognition system. Students will be able to chose their project from a list of projects. If there is a particular project you are interested in, please contact the TA directly. They will also be asked to present their project to the rest of the class and will be graded on the quality of the presentation as well as the project. There will be 2 presentations, an initial presentation and a final presentation on the project.
25% Homework -- There will be 3-4 homework assignments early in the term designed to provide students with the necessary skills to work on their projects.
15% Class Presentation -- Students will be asked to present a current paper from the object recognition literature. The grading will be based on how effectively the student conveys the information within the paper to the rest of the class.
Tu. Th 10:30-12:00