About this course
What does it mean to see the world, and how can a computer see? CSE 185 introduces the foundations of computer vision — from low-level image processing to modern deep learning models that recognize, segment, and reconstruct the visual world.
We will cover visual representations, projective geometry, image understanding, and object recognition, including topics like feature detection, image filtering, segmentation, motion analysis, 3D reconstruction (stereo, photometric stereo, structure from motion), and neural networks. Time permitting, we'll explore face recognition, pose estimation, detection & tracking, action recognition, transformer networks, and vision–language models.
Prerequisites
Required: CSE 031, CSE 100, MATH 024.
Strongly recommended: Linear algebra, multivariable calculus, probability/statistics, Python, and data structures & algorithms.
Course staff
How to reach us: use a public Piazza post for course questions so the whole class benefits, and a private Piazza post when the content should stay private (e.g., sharing a solution, a grading concern). Email is only for personal matters (accommodations, personal circumstances). For questions on course material or assignments, follow the chain Piazza → TA email → instructor email, giving reasonable time to respond at each step and linking the Piazza post (and any prior correspondence) each time you escalate. Not following this chain affects your participation grade.
Key dates
Assignment and midterm dates are tentative and will be confirmed in the first week.
8–11 AM, in-person
Weekly lab & office hours
| Day | Time | Activity |
|---|---|---|
| Tue | 4:30–6:20 PM | Lab Session |
| Tue | 6:20–7:20 PM | Office Hour (Kianna Ng, by appointment) |
| Wed | 7:30–8:30 AM | Office Hour (Maitrayee Keskar, by appointment) |
| Wed | 8:30 AM–12:20 PM | Lab Session |
| Wed | 12:20–1:20 PM | Office Hour (Maitrayee Keskar, by appointment) |
| Wed | 4:30–6:20 PM | Lab Session |
| Wed | 6:20–7:20 PM | Office Hour (Qhelile Sibanda, by appointment) |
Lecture schedule
| Mon | Topic | Wed | Topic |
|---|---|---|---|
| — | Aug 26 | Welcome & What is Computer Vision? | |
| Aug 31 | Image Types and Representations | Sep 2 | Geometric Image Formation |
| Sep 7 | No Class — Labor Day | Sep 9 | Camera Parameters |
| Sep 14 | No Class | Sep 16 | Midterm 1 Review |
| Sep 21 | Midterm 1 | Sep 23 | Acquiring Digital Images; Optics and Radiometry; Image Noise and Filtering |
| Sep 28 | Image Gradient — HW 1 due | Sep 30 | Edge Detection, Corner Detection |
| Oct 5 | Hough Transform, RANSAC, SIFT | Oct 7 | Midterm 2 — HW 2 due |
| Oct 12 | Camera Calibration | Oct 14 | Stereopsis 1 |
| Oct 19 | Stereopsis 2 | Oct 21 | Optical Flow |
| Oct 26 | Midterm 3 Review | Oct 28 | Midterm 3 — HW 3 due |
| Nov 2 | Structure from Motion 1 | Nov 4 | 3D Reconstruction |
| Nov 9 | Structure from Motion 2 | Nov 11 | No Class — Veterans Day |
| Nov 16 | Image Recognition and Object Detection | Nov 18 | Midterm 4 |
| Nov 23 | Template Matching | Nov 25 | No Class — Thanksgiving |
| Nov 30 | Convolutional Neural Networks 1 | Dec 2 | Convolutional Neural Networks 2 |
| Dec 7 | Vision Transformer | Dec 9 | Computer Vision for People |
| Dec 14 | Final Exam — 8:00–11:00 AM, in-person — HW 4 due | ||
How you'll be graded
Midterm clobber policy: your highest midterm score replaces your lowest — one free pass built in.
Grade scale
An A+ category, reserved for exceptional performance, is determined at the end of the semester.
Assignments
There are four assignments, one per course unit. Each combines the conceptual practice and the hands-on coding work for that unit into a single deliverable.
Working on them
Lab sessions are your built-in work time, with TAs on hand to help. Attendance isn't graded, but the labs are where most students do their best learning. If your section fills up or you want extra time, you're welcome to drop into additional sections.
You're encouraged to discuss concepts with classmates, but every solution you turn in must be written by you, in your own words.
Submitting your work
Assignments are submitted on Gradescope. Mark all pages associated with each problem — unmarked pages receive a 0 for that problem and will not be regraded.
Late work & slip days
Everyone gets 4 slip days to use across the four assignments. One slip day extends a deadline by 24 hours. They're meant for illness, submission issues, and other unforeseen circumstances — use them wisely and save them for when you actually need them. No need to ask — just submit late and we'll count it.
Once you're out of slip days, late work earns no credit — no exceptions. Staff actively prioritizes support for assignments whose deadlines haven't passed, so plan ahead. All work must be submitted by 11:59 PM on the final exam date; we won't accept submissions after that point, including slip days, so grades can go out on time.
Hitting a real obstacle (illness, family emergency, technical disaster)? Reach out early — the sooner we know, the more we can do.
Exams
Midterm quizzes (4)
- In-class, in-person — no alternative times.
- Your highest score replaces your lowest, effectively giving you a free quiz.
- Dates are TBD and will be announced in the first week.
Final exam
- Monday, December 14, 8:00–11:00 AM
- In-person, no alternate times.
- If you have a known conflict (e.g., religious observance, official UC event), report it before the first midterm.
- Bring a fully charged laptop unless told otherwise.
Policies
Academic integrity
Academic honesty is taken very seriously at UC Merced. Talk to each other about concepts — that's how you learn. But when it's time to turn in a solution, write it yourself.
Not allowed:
- Sharing or receiving solutions before or after a deadline.
- Submitting a friend's work to test the autograder.
- Examining someone else's solution for "ideas."
- Submitting ChatGPT / GitHub Copilot output as your own solution.
- Posting course solutions or materials publicly.
- Code obfuscation to dodge detection.
Penalties: first offense → the negative of the assignment's maximum points (and it cannot be dropped). Second offense → automatic F in the course. All cases are reported to the Office of Student Rights and Responsibilities.
Participation bonus
Up to 2.5% is added to your grade for active and appropriate use of Piazza, lab, and lecture to discuss course topics and ask & answer questions. Not following the guidelines below will reduce your participation score.
Getting help: use Piazza before emailing the instructor or TA, in the order Piazza → TA → Instructor, giving reasonable time for a response at each step. Emails about course material must include a link to the corresponding Piazza post. We'll do our best to respond to emails within 48 hours on weekdays.
Good participation looks like:
- Making Piazza posts public when you can — it lets classmates help (and lets others learn from the answer). Keep a post private only if it would reveal a solution others are still working on; when in doubt, go private. We may re-mark a post public or private, anonymously, if we think it'll help others.
- Searching before posting — your question may already be answered, and a redundant post affects your participation score.
- Asking specific questions. "How does computer vision work?" or "why is the solution this?" doesn't give us enough to work with — point to the exact lines or ideas that are confusing you.
- Coming to TAs with a test that isolates the bug and evidence you've stepped through it, rather than asking them to debug your code from scratch — part of this course is becoming a better tester and debugger yourself.
- Using the "good question" / "good answer" vote buttons instead of posting "+1" as an answer or follow-up.
Not part of participation:
- Filming, photographing, or recording class without permission.
- Asking about homework or upcoming labs during a lab section while classmates are still working on the current lab — be mindful of pacing so everyone gets the TAs' support.
Inclusion & classroom climate
You belong here. We welcome students from every background, identity, and path. If your name or pronouns differ from official records, let us know — we'll use what you prefer. If a religious holiday conflicts with an exam, tell us early and we'll work it out.
If something about the classroom environment isn't working for you, please bring it to us directly or by email. We can't fix what we don't hear about.
Extenuating circumstances
Extenuating circumstances are circumstances outside your control that directly inhibit your ability to complete assignments on time — things like unforeseen physical or mental health crises, technical issues, or family emergencies.
Tell us as soon as possible. The sooner we know, the more options we have to help. If you have a prolonged extenuating circumstance and expect to run out of slip days, reach out proactively before that happens.
Campus resources
- Counseling and Psychological Services — mental health support.
- Office for Prevention of Harassment and Discrimination — for survivors of sexual violence.
- 988 Suicide & Crisis Lifeline — call or text 988.
- Technology Resources Program — laptop and technology support.
- Dean of Students — (209) 228-3633.
- Student Accessibility Services — (209) 228-6996, access@ucmerced.edu.
Books & resources
There are many excellent computer vision textbooks, and a multitude of great resources online, so I hesitate to name one official textbook. Below is a list of books that cover topics in this course — I highly recommend reading the corresponding sections as topics come up (before class, after class, during class if you can manage it). I'll do my best to maintain an index of resources — textbook sections, articles, and videos — for each topic below. Exams aren't written from textbook trivia; learning the concepts from any of these books will serve you well.
Great books in computer vision
- Computer Vision: Algorithms and Applications — Szeliski (free online)
- Introductory Techniques for 3D Computer Vision — Trucco & Verri
- Computer Vision: A Modern Approach — Forsyth & Ponce
- Digital Image Processing — Gonzalez & Woods
- Multiple View Geometry — Hartley & Zisserman
- An Invitation to 3D Vision: From Images to Geometric Models — Ma, Soatto, Kosecka & Sastry
Supplemental textbooks & readings
- Deep and Shallow: Machine Learning in Music and Audio — Dubnov & Greer
- Fully Convolutional Networks for Semantic Segmentation
- U-Net: Convolutional Networks for Biomedical Image Segmentation
- Rich feature hierarchies for accurate object detection and semantic segmentation
- Edge Boxes: Locating Object Proposals from Edges
- You Only Look Once: Unified, Real-Time Object Detection
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Improved Deep Metric Learning with Multi-class N-pair Loss Objective
- Representation Learning with Contrastive Predictive Coding
- Contrastive Learning of Medical Visual Representations from Paired Images and Text
- Learning Transferable Visual Models From Natural Language Supervision
- Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
- Segment Anything
Recommended by topic
Letters correspond to author initials from the textbook list above (S = Szeliski, TV = Trucco & Verri, FP = Forsyth & Ponce, GW = Gonzalez & Woods, HZ = Hartley & Zisserman, DG = Dubnov & Greer).
- Image formation (Lec. 2 & 3): S Ch. 1 & 2; TV Ch. 2; FP Ch. 1 & 2; HZ 1.1–1.2, Ch. 2, 3.1–3.2, 6.1–6.2
- Spatial filtering (Lec. 4): GW Ch. 3
- Fundamental matrix, essential matrix, and epipolar geometry: HZ Ch. 9–11, TV Ch. 7
- Optical flow: S Ch. 9
- Video introduction to convolution
- 3Blue1Brown neural network series
- Hough Transform visualization by Isaac Sousa (CSE 185, 2025)
- Convolutional neural networks: S Ch. 5–6, DG Ch. 6.1–6.2, 7.3
FAQ
Click a question to expand.
How do I get help when I'm stuck?
Post on Piazza first — classmates and staff respond there, and your question probably helps others too. Come to office hours (Prof. Greer: Wed 2:45–3:45pm; TA hours by appointment — see the weekly schedule), or grab help during lab. Email is for private matters.
Can I use ChatGPT or Copilot?
Not as a solution generator. Pasting AI output as your assignment answer counts as an academic integrity violation. Using AI to learn concepts or debug your own code is fine — the line is what you submit.
What if I miss a midterm?
There are no make-up midterms. The good news: your highest midterm score replaces your lowest, so missing one is effectively your "drop." If something serious happens, contact Prof. Greer right away.
What if I have a conflict with the final exam?
Report it before the first midterm so we can plan ahead. The final is December 14, 8–11 AM. There are no alternate times unless your conflict was reported and approved early.
How do slip days work?
You get 4 slip days total to spend across the four assignments. Each one extends a deadline by 24 hours. Just submit late — we'll count them automatically. No request needed.
Can I collaborate on assignments?
You can — and should — discuss concepts with other students. But you must write your own solutions in your own words. If you can't explain it to me on the spot, you don't own the solution yet.
Do I have to come to lecture / lab?
Attendance isn't graded, but the lab is where most students cement what they learned in lecture. We strongly recommend showing up. Bonus participation points reward active engagement.
I need accommodations. What do I do?
Contact Student Accessibility Services ((209) 228-6996, access@ucmerced.edu) and forward your accommodation letter to Prof. Greer. Reach out as early as possible so we can make sure everything is in place.
I'm worried about my grade. Now what?
Come talk to us early — don't wait until week 14. Office hours exist for this. We can talk through study strategies, point you to resources, and help you figure out where the gap is.
Can I record lectures?
No filming, photographing, or audio recording of lectures without permission. If you need a recording for accessibility reasons, work through Student Accessibility Services.
Where will the schedule appear once dates are set?
On this page and on Piazza. We'll announce midterm and assignment dates in the first week of class.