Posts tagged cnn

MobileNet

Depthwise separable convolution explained: the building block that makes a CNN cheap enough to run on a phone.

• cnn
• computer-vision
• deep-learning
• Written in EN

Daniel Du
07 Apr 2023 · 3 min read

The Inception Network

How Google's Inception architecture stacks Inception modules into GoogLeNet, and what those auxiliary side branches are for.

• cnn
• computer-vision
• deep-learning
• Written in EN

Daniel Du
05 Apr 2023 · 3 min read

Inception Network Motivation

Why the Inception module saves you from choosing between 1x1, 3x3, 5x5 and pooling, and how 1x1 convolutions keep the cost down.

• cnn
• computer-vision
• deep-learning
• Written in EN

Daniel Du
03 Apr 2023 · 3 min read

A Simple Convolutional Network Example

A complete worked ConvNet for classifying 39x39 RGB images, layer by layer, from the first convolution down to the final softmax.

• cnn
• computer-vision
• deep-learning
• Written in EN

Daniel Du
27 Mar 2023 · 3 min read

Unleashing the Power of Convolutional Neural Networks

Why convolutional layers beat fully connected ones on images: parameter sharing, sparse connections, and 14 million parameters saved.

• cnn
• computer-vision
• deep-learning
• Written in EN

Daniel Du
07 Mar 2023 · 3 min read

Strided Convolutions

How stride changes the way a filter moves across an image, and the formula that gives you the resulting output size.

• cnn
• computer-vision
• deep-learning
• Written in EN

Daniel Du
27 Feb 2023 · 3 min read