Description
Prerequisites: You should have a graduate-level understanding of computer vision (equivalent of CS543 or CS445+CS444 combined) , including camera models, image filters, two-view geometry, feature detection and matching, and recognition. You should also have a graduate-level understanding of machine learning (equivalent of CS446). You should be engaged in or interested in research in 3D vision
Course Objective: Through this course, you will gain a broad and deep knowledge of the state-of-the-art in 3D vision, demonstrated and reinforced by group presentations and write-ups. You will also gain experience in identifying and validating research directions, as well as hands-on skills in implementing various 3D vision algorithms.
Requirements: This advanced-level course aims to help you master advanced techniques and gain insights into cutting-edge research topics. It assumes that students already have prerequisite knowledge; hence, it is not intended to teach fundamental knowledge or basic skills in computer vision. The discussions and lectures are heavily research-oriented and are better suited for graduate students in research programs related to areas such as vision, AR/VR, and robotics. Weekly readings, group presentations, coding assignments, write-ups, active class participation, and deliverables related to the final project are required.
Participation: We expect you to attend all lectures and discussions in-person unless pressing and unforeseen conflicts arise. Conflicts that are persistent (e.g. registering for two classes with half an hour overlap) are not excused.
Syllabus: Please see this this link.
Slack Sign-Up: Please sign up using this link.
Grading:- 35%: Role-play Discussion
- 25%: Non-Hacker Role-playing Presentations (5% each, tentative)
- 10%: Hacker’s Deliverables (code, demo, presentation)
- 40%: Final Project
- 10% Proposal (3-4 pages)
- 15% Poster & Demo Presentation
- 15% Final Report (6-8 pages)
- 10%: Research Topic Survey (4 pages)
- 10%: Quiz (2.5% each)
- 5%: Participation in class and on slack
