Computer Vision

Instructor

Course Description

Have you ever wondered how a machine or computer is made capable of understanding, interpreting, and giving semantics to an image/video? Have you ever thought about how image/video could be used to automate processes in a wider application domain ranging from industry to biomedicine? The answer lies in image processing and computer vision which plays a central role in all of these.

This course is a continuation of EE451 Digital Image Processing and is developed for Electrical Engineering and Computer Science students. Computer Vision has emerged as a revolutionary field that employs Image Processing, Pattern Recognition, and Machine Learning to imitate human vision in order to automate processes in a wide range of real-world applications.

The ever so rapid growth and hence increasing applications have made Computer vision ubiquitous and it, along with Artificial Intelligence, is transforming the way humans approach their daily routine life. In this course, the aim is to explore computer vision from both foundational and modern learning-based perspectives, covering fundamentals i.e. image formation/processing, deep learning for vision, 3D geometry/multi-view methods, motion estimation, and emerging areas such as radiance fields, generative models, and vision–language systems. The course emphasizes practical understanding through assignments and a semester-long project.

Course Aims

The course aims to equip students with:

• Build strong foundations in image formation, image processing, and feature extraction.
• Learn deep learning methods for vision, including neural networks, CNNs, transformers, and representation learning.
• Understand 3D vision and motion, covering camera geometry, multi-view reconstruction, and optical flow.
• Gain hands-on experience through programming assignments and a semester-long project.
• Develop the ability to read, implement, and present computer vision advances i.e. research and projects.

Week-Wise Schedule (Tentative)

Assignments

Textbooks & Materials

Assessments & Grading

Course Policies

Late Submission Policy

Assignments submitted after the deadline can still receive credit for up to 7 days (rounded up) after the original due time.

Late submissions are penalized using a grade multiplier that decreases linearly with each day late:

• On-time submissions receive full credit (multiplier = 1.0).
• 7 days late receives half credit (multiplier = 0.5).
• Submissions more than 7 days late receive no credit.

Late days are counted in whole days and rounded up. For example, if an assignment is due Tuesday at 11:59 PM, the final time to submit for partial credit is the following Tuesday at 11:59 PM.

If your assignment earns p points and is submitted t days late (where $0 \le t \le 7$), your final score is computed as:

$\left(1 - \frac{0.5t}{7}\right) p$

Importantly, the penalty depends only on the number of late days—not the hour of submission. Staying up late will not improve your score. If you're stuck, it's often better to rest, seek help, and continue the next day.

Free Late Days: Each student is granted 3 late days that can be used throughout the semester without any penalty. These are intended for special circumstances such as illness, travel, or other unexpected events. Late days are applied automatically to your first late submissions until exhausted. You do not need to inform us or request permission to use your free late days—simply submit when ready. However, free late days cannot be used to extend a submission beyond the 7-day maximum late window.

---

Policy on the Use of AI Assistants

You are welcome to use AI tools in this course in the same way you might use office hours or ask clarifying questions from an instructor. Appropriate AI uses include:

• Clarifying lecture concepts
• Understanding problem statements
• Getting high-level guidance or tips on how to get started

However, just as you would not ask another person to complete your assignment for you, you should not ask an AI to do your work or provide full solutions.

When in doubt, imagine the AI as a human helper and apply the same academic integrity standards.

If you use any AI assistance while working on a problem set, please include a short disclosure at the top of your submission describing:

• Which AI tool you used
• How you used it

A few sentences is sufficient.

Course Project

The course project accounts for 35% of the final course grade. The project is designed to evaluate your ability to define a problem, make steady technical progress, and clearly communicate your ideas and results.

It consists of three required components:

The proposal and progress report will be assessed by the course instructor.

Related Courses