Computer Vision
Use Deep Learning to build your own image classification model, and find out why the data you feed it might cause some unexpected results!
Overview
A terrible new virus has emerged and it’s making cats and dogs very ill. Fortunately, there are vaccines available and the government is rolling out an automated vaccination program using drones so that all stray cats and dogs can be targeted. However, the drone needs to use its on-board canera to identify the two species - otherwise they might get the wrong vaccine!
In this challenge, you will train a deep learning model to classify images. You will learn how Convolutional Neural Networks can be used to teach a computer how to distinguish between different objects, as well as learning about some of the common pitfalls we encounter when building these models - and how to avoid them.
Learning themes and outcomes:
- Understand how a classification model works, using training data within an algorithm to predict the label of instances not seen before.
- Understand the impact of biases in training data in prediction quality, and how it can impact our society.
- Describe what engineers can do to mitigate the effects of biased systems.
Getting Started
To get started, head to the Computer Vision challenge slides at this link.