CSCI 5561 Computer Vision (Spring 2025)

08:15 A.M. - 9:30 A.M., M & W Keller 3-111

Instructor: ​
​ Patrick Plonski [email protected] (Office hour: Monday and Wednesday 9:30am,
Keller 6-198. Please stop by after class or make an appointment).
TAs: ​
​ Qingyuan Jiang: Monday and Wednesday 16:00-17:00 (Shepherd Lab 161) and on
Fridays when homework is due.
Mason Hawver: Thursday and Friday 14:30-15:30 at Lind Hall L103.

Please use canvas for course related communication. In particular, we

will set up discussion boards for homework assignments and the
project so the entire class can benefit from the answers and the
discussion. If you need to reach the instructor or the TAs, please
include [CSCI 5561] at the beginning of your subject line.

Course Description
The primary goal of this course is to introduce students to computer vision -- how to make
computers make sense of images! Traditionally, computer vision has been studied in a bottom
up fashion. At the lowest level, images are processed to obtain simple features (e.g. corners,
edges). Then comes geometry in which these features are used to build 3D models of the
environment. Next up is mid-level vision in which features are grouped into segments or other
salient regions. At the highest level is the problem of object recognition which could be
considered as a gateway to general intelligence. In the last decade, the field has been
transforming at a mind boggling rate where learning based architectures have become
increasingly dominant. It is impossible to cover all aspects of computer vision in depth in a
single course. We will focus on fundamentals and let you explore state of the art methods
through the course project The course will be organized around three modules: Geometry,
Image Features and Deep Learning. After taking this course, you will:

●​ Learn about basic tools and techniques for designing (and analyzing) computer vision
algorithms and architectures
●​ Become familiar with active research challenges in vision
●​ Gain hands-on experience in building vision systems

Important Note about Required Background -- Please read carefully!

While this is an introductory course in computer vision, anyone who is interested in taking it
needs to understand that it is a graduate level computer science course!
This means that you need a certain degree of mathematics and algorithms background in order
to take this course. In particular:

Prerequisites: CSci 5521 Machine Learning is a required prerequisite. This requirement will be
enforced. These prerequisites are there to ensure that you have the sufficient background in
●​ Linear algebra
●​ Probability and statistics
●​ Optimization, including algorithmic methods such as dynamic programming
●​ Programming: in-class exercises and projects are in Python
When discussing vision algorithms, we will assume that you are familiar with the material from
CSCI 5521.

Topics Covered and Text Book

The textbook for the course is Foundations of Computer Vision by Torralba, Isola and
Freeman. The book is available through the library. A link is provided in Canvas. It is a new
textbook which aligns well with the way this class is taught. In particular, we will cover
●​ Module I: Geometry: Section XI (Ch. 38-44) and Section XII (Ch. 46-49)
●​ Module II: Features: Section IV (Ch. 15-16), Section V (Ch. 17-19). Section VI (Ch.
20-23) + SIFT
●​ Module III: Neural Architectures Section VII (Ch. 24-26) + Classification, Detection and
Segmentation from Understanding Deep Learning (see below for reference)

There are many excellent resources online, and feel free to use them. For classical textbooks,
●​ For signal and image processing, there are classic textbooks such as​
Oppenheim’s Signals and Systems​
Gonzales and Wood’s Digital Image Processing are standard references. ​
Another good resource is Hany Farid’s manuscript.​
You might want to refer to these books when we cover concepts such as convolution,
filtering and sampling (which are also the foundations of all convolutional neural network
architectures)
●​ Understanding Deep Learning by Simon Prince is a great resource for the deep learning
component
●​ Multiple View Geometry in Computer Vision by Hartley and Zisserman is the standard
reference for geometry.
●​ For general computer vision texts, the two recommended textbooks are Computer
Vision: A Modern Approach by Forsyth and Ponce (2nd Edition) , and Rick Szeliski's
book.
●​ Trucco & Verri’s Introductory Techniques for 3-D Computer Vision is a classic
●​ Prof Park’s slides
●​ I will make my lecture notes available
Expectations
We will cover a relatively large range of topics which rely on various background material. As a
result, at times, you might find that the current topic is quite challenging (or rather basic). For
example, dynamic programming may be easy to grasp for a student who has taken advanced
algorithms courses. The same student may find projective geometry quite challenging. Here are
some tips that you might helpful:

●​ Do not rely on a single source to learn the material. If the topic in one book does not
make sense, do not hesitate to review other books, or even basic texts such as your
linear algebra textbook. If you need guidance, do not hesitate to talk to your instructor or
the TA. There are excellent resources online.
●​ Learn from each other.
●​ Work regularly. As the list of topics is quite diverse, make sure that you review the
material in a timely fashion.

Evaluation
Evaluation will be composed of the following components:

●​ Homework (45%): Up to five assignments. Important: The assignments will require

writing (lots of!) python code.
●​ Midterm (10%): You need to score at least 50% in the midterm to pass this class.
Make-ups will be given only if there are extenuating circumstances proven with a valid
excuse letter. The make-up exam could be an oral exam.
●​ Course Project (45%): More about the project will be posted on the project page.

Grades: 93.0% or above yields an A, 90.0% A-, 86% = B+, 82% = B, 78% = B-, 74% = C+, 70%
= C, 67% = C-, 63% = D+, 60% = D, and less than 60% yields an F.

Other grade related issues: We will give ample time for each homework assignment. It is critical
that you start early. The assignments will be due Fridays. If you submit late, but before the
following Monday, you get 75% of the grade. Afterwards, 0%. Questions about a specific grade
should be raised within a week after the grade is given. Incompletes will, in general, not be
given. Exceptions will be considered only when a provably serious family or personal
emergency arises, proof is presented, and the student has already completed all but a small
portion of the work.

Scholastic Misconduct
Scholastic misconduct is broadly defined as "any act that violates the right of another student in
academic work or that involves misrepresentation of your own work. Scholastic dishonesty
includes, (but is not necessarily limited to): cheating on assignments or examinations;
plagiarizing, which means misrepresenting as your own work any part of work done by another--
including generative AI tools; submitting the same paper, or substantially similar papers, to meet
the requirements of more than one course without the approval and consent of all instructors
concerned; depriving another student of necessary course materials; or interfering with another
student's work."

Important: Computer Vision libraries should make your life easy but should not do the
homework for you! You are welcome to use standard matrix functions, or utilities to read and
write images, but when, e.g., the problem asks you to implement an edge detector, you should
not use built-in edge detection functions. When in doubt, just ask. Using implementations
off-the-web or elsewhere in homeworks yields an immediate F. If you use code from anyone
or anywhere, you must give reference. Otherwise, this constitutes cheating. While it is
allowed to discuss homework with other students at a high level, you should write your own
code independently. Do NOT use other students’ code ( including students who have taken this
class previously). An automatic plagiarism system will be run to determine the similarity of code.
All plagiarism will be reported to the Dean’s Office and the Office of Community Standards.

A Note from DRC

The University of Minnesota is committed to providing equitable access to learning opportunities

for all students. Disability Resource Center (DRC) is the campus office that collaborates with
students who have disabilities to provide and/or arrange reasonable accommodations.

●​ If you have, or think you may have, a disability (e.g., mental health, attentional, learning,
chronic health, sensory, or physical), please contact DRC at 612-626-1333 to arrange a
confidential discussion regarding equitable access and reasonable accommodations.
●​ If you are registered with DRC and have a current letter requesting reasonable
accommodations, we encourage you to contact Prof. Isler early in the semester to review
how the accommodations will be applied in the course.

Additional information is available on the DRC website: https://diversity.umn.edu/disability/

As a student you may experience a range of issues that can cause barriers to learning, such as
strained relationships, increased anxiety. alcohol/drug problems, feeling down, difficulty
concentrating and/or lack of motivation. These mental health concerns or stressful events may
lead to diminished academic performance or reduce your ability to participate in daily activities.
University of Minnesota services are available to assist you with addressing these and other
concerns you may be experiencing. You can learn more about the broad range of confidential
mental health services available on campus via http://www.mentalhealth.umn.edu

If you require disability or other accommodations, please reach out to discuss them as soon as
possible.