This repository includes the PDF slides and LaTeX source for the Computer Vision track at Stanford's 2019 AI4ALL camp. The curriculum aims to introduce mathematical concepts alongside visual intuitions, beginning with linear regression and vector spaces and ending with deep convolutional neural networks. It assumes no prior understanding apart from having seen multivariable linear equations, but some basic understanding of calculus can be helpful. The entire slide deck covers around 4-6 hours of material, depending on the amount of time devoted to group discussion and questions.
This material borrows from Stanford's CS231N course. It is notably modified with introductory slides for relevant concepts, a heavier reliance on visual intuition, and more student interaction.
The following is the section break down for each lecture.
-
Linear Models for Regression and Classification
-
Optimization
-
Vectors
-
Matrices
-
Image Representation
-
Image Classification
-
Challenges of Computer Vision
-
Neural Networks
-
Recap: Image Representation
-
Recap: Challenges of Computer Vision
-
Recap: Neural Networks
-
Invariant Local Features
-
Classical Feature Extraction
-
Deep Convolutional Neural Networks