📗 -> Computer Vision
short mode
not done
tags include #class
description includes CV🔗 Critical Links
- Course Website
- Textbook 1: Computer Vision - A modern approach 2nd ed. by Forsyth and Ponce
- Online Textbook 2
🔶 Course Description
Computer vision is the study of enabling machines to “see” the visual world; e.g., understand images and videos. Explores several fundamental topics in the area, including feature detection, grouping and segmentation, and recognition.
❗ Important
Instructor:
Hamed Pirsiavash
hpirsiav@ucdavis.edu
Office hours: Thursdays at 4pm-5pm on Zoom or by appointment.
TA:
Kossar Pourahmadi-Meibodi
kmeibodi@ucdavis.edu
Office hours: Mondays and Wednesdays at 4pm-5pm in Kemper 55 (starting on Oct 7).
Raymond Kang
rhkang@ucdavis.edu
Office hours: Tuesdays at 4pm-5pm in Kemper 55 (starting on Oct 1).
Final Project
40 Points
• Big Project– Better if related to your own research
• Proposal is due on 4/30/2023 at 11:59pm
– Less than a page
– The problem that you want to solve
– Motivation: why is this an important problem to solve.
– What dataset you will use.
– What method you will use to solve it.
– Crisp final outcome/deliverable
• Final report, code, and presentation video due on 6/7/2023 at 11:59pm
ECS174-Final-Project
ECS174-Final-Prep
📄 Class Material
Week 1 - Introduction
- Introduction: ECS174-L1
- A general introduction to applications and issues in CV
Week 2 - Images and Filters / Image Sampling
Week 3 - Edges / Image Transformations
-
Review and Sampling Continued: ECS174-L4
-
Edges: ECS174-L5
-
Image Transformations
-
Discussion - Resource Folder
Week 4 - Interest Points / Descriptors
- Interest Points
- Jump ahead to learning…
- Descriptors
- Skipped lecture to study for midterm
- ECS174-L7
- Discussion - Resource Folder
Week 5 - Image Stitching / Learning
- Image Stitching
- Learning
Week 6 - Neural Networks / Invited Speaker or TA
- Neural Networks
- Invited Speaker or TA (Hamed will travel for ICLR conference)
Week 7 - Invited Speaker or TA / Convolutional Neural Networks
- Invited Speaker or TA (Hamed will travel for ICLR conference)
- Convolutional Neural Networks
Week 8 - Convolutional Neural Networks / Optical Flow and Motion
- Convolutional Neural Networks
- Optical Flow and Motion
Week 9 - Stereo / More on CNNs
- Stereo
- More on CNNs
Week 10 - Object Detection / Thanksgiving Holiday
- Object Detection
- Thanksgiving Holiday (no class)
Week 11 - Self-supervised Learning / Generative Adversarial Networks (GAN)
- Self-supervised Learning
- Generative Adversarial Networks (GAN)
https://web.cs.ucdavis.edu/~hpirsiav/courses/CVf24/slides/8_lecture_13_SSL.pdf- Contrastive learning very effective
Masked Autoencoders Are Scalable Vision Learners slideshow
- Contrastive learning very effective