Introduction

Completed

Image classification is a common workload in artificial intelligence (AI) applications. It harnesses the predictive power of machine learning to enable AI systems to identify real-world items based on images.

Uses of image classification

Some potential uses for image classification include:

  • Product identification: performing visual searches for specific products in online searches or even, in-store using a mobile device.
  • Disaster investigation: identifying key infrastructure for major disaster preparation efforts. For example, identifying bridges and roads in aerial images can help disaster relief teams plan ahead in regions that are not well mapped.
  • Medical diagnosis: evaluating images from X-ray or MRI devices could quickly classify specific issues found as cancerous tumors, or many other medical conditions related to medical imaging diagnosis.

Learning objectives

Learn how to use the Custom Vision service to create an image classification solution. In this module you will:

  • Identify image classification scenarios and technologies
  • Provision a Custom Vision resource and use the Custom Vision portal
  • Train an image classification model
  • Publish and consume an image classification model