Computer Vision with GluonCV

In this course, you will build a useful understanding of the components of a convolutional neural network (CNN) like convolutions and pooling layers etc. In this course, Alex Smola and Tong He show how to implement some computer vision techniques using GluonCV, a computer vision toolkit.

Intended Audience

This course is intended for:

• Developers who are looking to implement common computer vision models

Course Objectives

In this course, you will learn how to:

• Summarize various convolutional neural network components like convolutions, padding, and channels
• Translate the components to code when creating a neural network like LeNet
• Import your data into a Gluon Data Loader for training and transformation

Prerequisites

We recommend that attendees of this course have the following prerequisites:

• A basic understanding of artificial neural networks
• A basic understanding of linear Algebra topics like matrices, matrix multiplication, and dot products

Delivery Method

This course is delivered through:

• Digital training

Duration

• 2 hours

Course Outline

This course covers the following concepts:

• Convolutions
• Padding and stride
• Channels
• Pooling
• LeNet
• Activation functions
• DropOut
• Batch normalization
• Blocks
• The curse of the last layer
• Residual networks
• Data processing