Carnegie Mellon University: AI & Vision Courses Reference¶
A reference of CMU's AI, machine learning, and computer vision courses relevant to the AI Hardware Engineer Roadmap — compiled from official course pages and catalogs.
Sources: - 07-280 AI & ML I - 15-463 Computational Photography - 16-385 Computer Vision Spring 2026 — Lectures - Szeliski: Computer Vision (2nd ed.)
Table of Contents¶
- 07-280: AI & Machine Learning I
- 15-463: Computational Photography
- Graphics & Imaging Courses
- Self-Study Mapping to Roadmap
- Additional Resources: Szeliski Book & Related Courses
1. 07-280: AI & Machine Learning I¶
New in Spring 2026 — Replaces 15-281 and 10-315. Foundation for 07-380 AI & ML II.
Overview¶
| Item | Details |
|---|---|
| Lectures | Tue + Thu, 11:00 am–12:20 pm, Tepper 1403 |
| Recitation | Friday afternoon (5 sections) |
| Instructors | Nihar Shah, Pat Virtue |
| Education Associate | Brynn Edmunds |
| Textbook | No required textbook; readings from AIMA, Bishop, Daume, Goodfellow, MML, Mitchell, Murphy, KMPA (all online or via CMU Library) |
Course Description¶
Integrated introduction to AI and ML bridging core methods with modern approaches. Students build implementations of landmark systems: AlexNet, GPT-2, and AlphaZero. Covers ethics and responsible AI development.
Grading¶
| Component | Weight |
|---|---|
| Midterm 1 | 15% |
| Midterm 2 | 15% |
| Final Exam | 25% |
| Programming/Written Homework | 30% |
| Online Homework | 5% |
| Pre-reading Checkpoints | 5% |
| Participation (in-class polls) | 5% |
Grade boundaries (rough): A >=90%, B 80-90%, C 70-80%, D 60-70%. Not curved.
Prerequisites (Strict)¶
- 15-122 Principles of Imperative Computation
- Probability (concurrent)
- Linear Algebra (prior)
- 15-151 or 21-127 Mathematical Foundations (prior)
- Calculus 2 (concurrent)
Schedule (Spring 2026)¶
| Dates | Topic |
|---|---|
| 1/13 | 1. Introduction |
| 1/15 | 2. Heuristic Search |
| 1/20 | 3. Adversarial Search |
| 1/22 | 4. Constraint Satisfaction Problems |
| 1/27 | 5. ML Problem Formulation |
| 1/29 | 6. Decision Trees |
| 2/3 | 7. Linear Regression |
| 2/5 | 8. Optimization |
| 2/10 | 9. Logistic Regression |
| 2/12 | 10. Feature Engineering and Regularization |
| 2/17 | 11. Neural Networks |
| 2/19 | 12. Neural Networks (cont.) |
| 2/24 | Midterm Exam 1 |
| 2/26 | 13. AI Alignment |
| 3/3, 3/5 | Spring Break |
| 3/10 | 14. PyTorch, Autograd, Pre-training/Transfer/Fine-tuning |
| 3/12 | 15. Deep Learning for Computer Vision, GPUs |
| 3/17 | 16. MLE and Probabilistic Modeling |
| 3/19 | 17. NLP, Markov Chains, N-grams |
| 3/24 | 18. Feature Learning, Word Embeddings |
| 3/26 | 19. NLP: Attention, Position Encoding |
| 3/31 | 20. Transformers, LLMs |
| 4/2 | 21. Markov Decision Processes |
| 4/7 | 22. Reinforcement Learning |
| 4/9 | Carnival (no class) |
| 4/14 | 23. Deep Reinforcement Learning |
| 4/16 | 24. Monte Carlo Tree Search |
| 4/21 | Midterm Exam 2 |
| 4/23 | 25. AI/ML Ethics |
Assignments¶
| HW | Type | Due |
|---|---|---|
| HW0 | Online | 1/15 Thu |
| HW1 | Online, Written, Programming | 1/22 Thu |
| HW2 | Online, Written, Programming (Search & Games) | 1/29 Thu |
| HW3 | Online only | 2/5 Thu |
| HW4 | Written only | 2/12 Thu |
| HW5 | Mostly Programming | 2/19 Thu |
| HW6 | Mostly Written | 2/26 Thu |
| HW7 | Mostly Programming | 3/12 Thu |
| HW8 | Building AlexNet | 3/19 Thu |
| HW9 | Online, Written, Programming | 3/26 Thu |
| HW10 | Online, Written, Programming | 4/2 Thu |
| HW11 | Building GPT-2 | 4/16 Thu |
| HW12 | Building AlphaZero | 4/23 Thu |
Policies¶
- Late days: 6 total across all assignments; max 2 per assignment
- Pre-reading: Lowest 2 checkpoints dropped
- Participation: >=80% of in-class polls for full credit
- Collaboration: Conceptual discussion allowed; no sharing code/text; generative AI may not be used to generate submissions
- Programming partners: Groups of 2 allowed for programming components only
Comparison: 07-280 vs 10-301¶
| 07-280 | 10-301 |
|---|---|
| Heuristic Search, Adversarial Search, CSPs | — |
| ML Parallelism/GPU Basics | — |
| Monte Carlo Tree Search | — |
| Transformer networks, LLMs | Y |
| Reinforcement Learning | Y |
| ML fundamentals (decision trees -> neural nets) | Y |
| Fulfills AI Major core | Y |
| Prereq for 07-380 AI & ML II | Y |
2. 15-463: Computational Photography — Deep Insight Before Computer Vision¶
Recommended before 16-385 Computer Vision. Provides foundational understanding of imaging physics, camera pipelines, and computational methods that underpin both graphics and vision.
Source: 15-463 Fall 2018
Overview¶
| Item | Details |
|---|---|
| Cross-listing | 15-463 (undergrad), 15-663 (Master's), 15-862 (PhD) |
| Schedule | Mon + Wed, 12:00-1:20 PM |
| Textbook | Computer Vision: Algorithms and Applications (Szeliski), free online |
Course Description¶
Computational photography is the convergence of computer graphics, computer vision, and imaging. It overcomes traditional camera limitations by combining imaging and computation for new ways of capturing, representing, and interacting with the physical world.
Topics: modern image processing pipelines (mobile/DSLR), image/video editing, 3D scanning, coded photography, lightfield imaging, time-of-flight, VR/AR displays, computational light transport. Advanced topics: cameras at light speed, non-line-of-sight imaging, seeing through tissue.
Prerequisites (one of)¶
- 18-793 Image and Video Processing
- 15-462 Computer Graphics, OR
- 16-720 Computer Vision, OR
- 16-385 Computer Vision
Linear algebra, calculus, programming, and image computation required.
Grading¶
| Component | Weight |
|---|---|
| 7 Homework Assignments | 70% |
| Final Project | 25% |
| Class Participation | 5% |
Late policy: 6 free late days total; each additional late day = 10% penalty; max 4 days late per assignment.
Syllabus (Fall 2018)¶
| Topic |
|---|
| Introduction |
| Digital photography pipeline |
| Pinholes and lenses |
| Photographic optics and exposure |
| High dynamic range imaging |
| Tonemapping and bilateral filtering |
| Color |
| Image compositing |
| Gradient-domain image processing |
| Focal stacks and lightfields |
| Deconvolution |
| Camera models and calibration |
| Two-view geometry |
| Radiometry and reflectance |
| Photometric stereo |
| Light transport matrices |
| Computational light transport |
| Stereo and structured light |
| Time-of-flight imaging |
| Non-line-of-sight imaging |
| Fourier optics |
| Monte Carlo rendering 101 |
Assignments¶
7 homework assignments with programming (Matlab) and photography (DSLR) components. Final project may use lightfield cameras, ToF cameras, depth sensors, structured light systems.
Why Before Computer Vision¶
15-463 builds intuition for how images are formed (optics, radiometry, sensors) and how to process them (HDR, deconvolution, calibration, stereo). This physical and algorithmic foundation makes 16-385 Computer Vision (detection, recognition, geometry) much easier to grasp.
3. Graphics & Imaging Courses¶
| Code | Name | Notes |
|---|---|---|
| 15-462 | Computer Graphics | Rendering, transforms, shading — foundation for 15-463 |
| 15-463 | Computational Photography | Imaging physics, camera pipelines, HDR, lightfields — deep insight before 16-385 |
| 16-385 | Computer Vision | Image processing, detection, recognition, geometry-based vision |
4. Self-Study Mapping to Roadmap¶
For learners following the AI Hardware Engineer Roadmap, here is how CMU's AI/vision curriculum aligns:
| Roadmap Phase | CMU Course(s) | Overlap |
|---|---|---|
| Phase 3: Neural Networks | 07-280 (Neural Nets, AlexNet, PyTorch) | Graphs, training, autodiff before Phase 4 Track A/B hardware |
| Phase 3: Computer Vision | 15-463 (Computational Photography) -> 16-385 (Computer Vision) | Imaging physics, camera pipelines, then detection/recognition |
| Phase 3: Edge AI | 07-280 (deployment themes), Jetson-adjacent labs | On-device pipeline, latency/privacy context; pairs with Phase 4 Track B |
| Phase 3: Sensor Fusion | 15-463 (imaging), 16-385 (Vision) | Multi-sensor perception before Phase 4 Jetson integration |
| Phase 4 Track B (Jetson) | 07-280, 16-385, Jetson/Holoscan labs | Models on device, pipelines, latency |
| Phase 5: Edge Computing | 07-280, 16-385, Jetson/Holoscan | Efficient models, streaming, Holoscan |
| Phase 5: AI Chip Design | 07-280 (optimization, GPU/parallel), 16-211 (math) | Parallel compute, linear algebra for accelerators |
Suggested Self-Study Order (CMU-Inspired, AI focus)¶
- 15-122 equivalent — Imperative programming (C/Python)
- 21-120, 21-122, 21-241 — Calculus, linear algebra
- 07-280 topics — Search -> ML -> Neural Nets -> RL -> Transformers
- 15-462 (optional) — Computer Graphics — rendering, transforms
- 15-463 — Computational Photography — imaging physics, camera pipelines, HDR, lightfields (deep insight before vision)
- 16-385 — Computer vision
5. Additional Resources: Szeliski Book & Related Courses¶
Computer Vision: Algorithms and Applications (2nd ed.)¶
https://szeliski.org/Book/ — Richard Szeliski, University of Washington (c 2022)
The canonical computer vision textbook. Free PDF download for personal use. Used by 15-463 Computational Photography and many vision courses worldwide. Covers image formation, feature detection, stereo, structure from motion, recognition, and more.
Related Courses (from Szeliski Book)¶
Additional good sources for computer vision and computational photography, sorted roughly by most recent:
| Course | Institution | Instructor(s) | Term |
|---|---|---|---|
| CS5670 Introduction to Computer Vision | Cornell Tech | Noah Snavely | Spring 2025 |
| 6.8300/6.8301 Advances in Computer Vision | MIT | Bill Freeman, Antonio Torralba, Phillip Isola | Spring 2023 |
| 16-385 Computer Vision | CMU | Matthew O'Toole | Fall 2024 |
| 16-385 Computer Vision — Lectures | CMU | — | Spring 2026 |
| CS194-26/294-26 Intro to Computer Vision and Computational Photography | Berkeley | Alyosha Efros | Fall 2024 |
| 15-463, 15-663, 15-862 Computational Photography | CMU | Ioannis Gkioulekas | Fall 2024 |
| CSCI 1430 Computer Vision | Brown | James Tompkin | Spring 2025 |
| CMPT 412 and 762 Computer Vision | Simon Fraser | Yasutaka Furukawa | Fall 2023 |
| CS 4476-A / 6476-A Computer Vision | Georgia Tech | James Hays | Fall 2022 |
| EECS 498.008 / 598.008 Deep Learning for Computer Vision | U Michigan | Justin Johnson | Winter 2022 — outstanding intro to deep learning and visual recognition |
| DS-GA 1008 Deep Learning | NYU | Yann LeCun, Alfredo Canziani | Spring 2021 |
| Fundamentals and Trends in Vision and Image Processing | IMPA | Luiz Velho | Spring 2021 |
| CS294-158 Deep Unsupervised Learning | UC Berkeley | — | Spring 2020 |
| CSCI 497P/597P Introduction to Computer Vision | Western Washington | Scott Wehrwein | Spring 2020 |
| EECS 504 Foundations of Computer Vision | U Michigan | Andrew Owens | Winter 2020 |
Course links are maintained at szeliski.org/Book. Contact the author to add your course.
Links¶
| Resource | URL |
|---|---|
| 07-280 AI & ML I | https://www.cs.cmu.edu/~07280/#schedule |
| 16-385 Computer Vision Spring 2026 Lectures | https://16385.courses.cs.cmu.edu/spring2026/lectures |
| 15-463 Computational Photography | https://graphics.cs.cmu.edu/courses/15-463/2018_fall/ |
| Szeliski: Computer Vision (2nd ed.) | https://szeliski.org/Book/ |
Last updated: February 2025. Course offerings and requirements may change; verify with CMU official sources.