CSE 795/895: Computer Vision

Course Description

Computer Vision techniques serve as the foundation for many leading modern AI technologies.

Research interest in Computer Vision is at an all time high, due to significant advances in deep learning methods and computing harware as well as the availability of large and diverse datasets.

This course will discuss fundamental concepts of Computer Vision, the state-of-the-art tools and techniques, as well as advanced CV research topics.

At the end of the course, you will be familiar with many computer vision tasks and applications. You will also have gained hands-on experience via an exciting CV project.

Office Hours

Friday: 2:00 pm - 3:00 pm and by appointment

Topics

Textbooks

There is no official textbook for the course. The course content will be based on lecture slides and research papers published in the premier conferences such as CVPR and ICCV.

Grading Criteria

Attendance

Optional. Lectures will be screen-recorded and posted on Canvas

Honor Code

Please refer to the statement on academic integrity given below.

By attending Old Dominion University you have accepted the responsibility to abide by the honor code. If you are uncertain about how the honor code applies to any course activity, you should request clarification from the instructor. The honor code is as follows:

“I pledge to support the honor system of Old Dominion University. I will refrain from any form of academic dishonesty or deception, such as cheating or plagiarism. I am aware that as a member if the academic community, it is my responsibility to turn in all suspected violators of the honor system. I will report to Honor Council hearings if summoned.”

In particular, submitting anything that is not your own work without proper attribution (giving credit to the original author) is plagiarism and is considered to be an honor code violation. It is not acceptable to copy source code or written work from any other source (including other students), unless explicitly allowed in the assignment statement. In cases where using resources such as the Internet is allowed, proper attribution must be given.

Any evidence of an honor code violation (cheating) will result in a 0 grade for the assignment/exam, and the incident will be submitted to the Department of Computer Science for further review. Evidence of cheating may include a student being unable to satisfactorily answer questions asked by the instructor about a submitted solution. Cheating includes not only receiving unauthorized assistance, but also giving unauthorized assistance.

Students may still provide legitimate assistance to one another. You are encouraged to form study groups to discuss course topics. Students should avoid discussions of solutions to ongoing assignments and should not, under any circumstances, show or share code solutions for an ongoing assignment.

Please see the ODU Honor Council’s webpage for other concrete examples of what constitutes cheating, plagiarism, and unauthorized collaboration. All students are responsible for knowing the rules. If you are unclear about whether a certain activity is allowed or not, please contact the instructor.

Content

Lecture 1

Lecture 2

Lecture 3

Lecture 4

Lecture 5

Lecture 6

Lecture 7

Lecture 8