Using Computer Vision to Streamline Accessibility

Images are a critical part of websites. They convey meaningful ideas fast and even improve the retention rate. But not all users can benefit from images. According to a 2015 study from the American Foundation for the Blind, more than 23.7 million Americans have some form of vision loss.

Drupal is known for its accessibility. Content managers can easily add alternative text to images that visually impaired users can read instead of seeing the image. Unfortunately, adding this metadata is a step that is often skipped because it’s viewed as tedious or sometimes, even perceived as unimportant. In this session, we will show how we integrated Drupal with computer vision APIs to do this for you.

You'll learn how we worked with Google Vision API module that feeds an uploaded image into a visual recognition system (Google's Cloud Vision API) to get back a textual description of the image, which can be used as the image’s alternative text. We'll walk through the code, do a couple of demos, and speak a bit about the future of leveraging AI to make a smarter CMS.

In this session you'll learn:

  • Why alternative text is important to users and to SEO

  • How Computer Vision can automate the process of describing images

  • How to configure and test the Google Vision API module

  • What is next for Computer Vision and Drupal

Session Track

Content and Editorial

Experience level

Intermediate

Drupal Version

Drupal 8 (current)