CS C280, Computer Vision (Spring 2025)
Logistics
UC Berkeley, Spring 2025
Time: Monday, Wednesday 2:00PM - 3:29PM
Location: 1102 Berkeley Way West
Ed: CS280 SP25 Ed
Gradescope Code: 8KVE8D
Instructors: Jitendra Malik, Angjoo Kanazawa
GSIs: Jathushan Rajasegaran, Rahul Ravishankar, Ryan Tabrizi
Office hours - Location: TBD
- Jitendra Malik: After lecture
- Angjoo Kanazawa: After lecture
- Jathushan: TBD
- Rahul: TBD
- Ryan: TBD
Policy: Please see the syllabus (not yet available) for administrative details. To keep discussions organized, please do not email the instructor or GSIs directly. We are using Ed Discussion for communication (private or public) as much as possible.
Prerequisites: Students should have mastery over content from COMPSCI 189, COMPSCI 182, COMPSCI 180/280A. We will not be teaching basic image processing (convolution, gaussian smoothing, fourier transforms), basic deep learning (PyTorch, Jax, TensorFlow), stereo/homography.
Lectures
Lecture slides will be posted after leture.
- Lecture 1: (1/22) Course Intro. and Fundamentals of Image Formation
- Lecture 2: (1/27) Camera Calibration
- Lecture 3: (1/29) Calibration, Binocular Perspective
- Lecture 4: (2/03) Multi-View Geometry
- Lecture 5: (2/05) Dynamic Perspective
- Lecture 6: (2/10) Radiometry of Images
- Lecture 7: (2/12) Color Vision
- Lecture 8: (2/19) Human Vision
- Lecture 9: (2/24) Sequence Models
- Lecture 10: (2/26) ViT and Self-Supervised Models
- Lecture 11: (3/03) Diffusion Models
- Lecture 12: (3/05) Diffusion Models 2: Practicalities
- Lecture 13: (3/10) More Generative Models (VAEs, GANs, Evaluation Metrics)
- Lecture 14: (3/12) From pixels to objects
- Lecture 15: (3/17) Early History of ANNs; Object Detection; Face Recognition
Assignments
- HW0: Due Monday, 01/27 at 11:59pm
- HW1: Due Tuesday, 02/11 at 11:59pm
- HW2: Due Saturday 03/15 at 11:59pm