This course covers visual recognition, i.e., recognition methods applied to images, videos and other forms of visual data. Topics include object categorization, semantic segmentation, boundary detection, face recognition, human pose estimation and action recognition. The course will focus predominantly (but not exclusively) on deep learning which has recently emerged as one of the most prominent and successful approaches to tackle hard recognition problems. Deep learning is a special form of machine learning where rich data representations are simultaneously learned with the model, thus eliminating the need to engineer features by hand.
This class is based on reading, presentation and discussion of recently published papers in this area. Students will also be required to propose and complete a term project in the area of visual recognition. There is no midterm or final exam. Enrollment is by instructor's permission. Send us an email to be authorized to register.
- Lorenzo Torresani | 109 Sudikoff | office hour: by appointment
- Tue&Th 10-11:50am | x-hour (used occasionally to make up cancelled classes) W 3:00-3:50pm
Location: Sudikoff 213
Grading and policies
In order to encourage independent critiquing and to foster in-class discussion, you will NOT be permitted to talk about the assigned papers with your classmates before the lecture: I am interested in hearing your own personal views of the articles and not consensus opinions emerged from group discussions held before the lecture.
You are allowed to use external software for portions of your project. However, you should clearly report the use of external code and include pointers to such software in your project write-up. The project grade will be based on the novelty of your solution/application but also on the amount of new code written by you to implement the idea. So keep this in mind when considering to use software written by someone else.
These rules will be strictly enforced and any violation will be treated seriously