Project

Webpage

CS231n: Deep Learning for Computer Vision

Videos

Stanford University CS231n, Spring 2017

Summary

1. global average pooling: average over different feature maps H*W average pooling: average over pixels

Lec2 Image Classification Pipeline

K-Nearest Neighbor Classifier(KNN)

‣

Lec3 Loss function and optimization

Linear Classification

Loss Function & Regularization

Optimization

Lec4 Neural Networks and Backpropagation

Neural Network

Activation Functions, Weight Initialization & Normalization

Backpropagation: to compute gradient