# Alternate text and automation

DevFeed: [Alternate text and automation](<https://devfeed.tech/articles/alternate-text-and-automation-9368.md>)

Original publisher: [Read original article](<https://a11yproject.com/posts/alternate-text-and-automation/>)

Author: Eric Bailey

Published: 2021-03-23T00:00:00Z

Content type: article

Language: en

Sources: [The A11Y Project](<https://devfeed.tech/sources/the-a11y-project.md>)

Topics: [alt text](<https://devfeed.tech/topics/alt-text.md>), [Accessibility](<https://devfeed.tech/topics/accessibility.md>), [Web Content Accessibility Guidelines](<https://devfeed.tech/topics/web-content-accessibility-guidelines.md>), [Content Management System](<https://devfeed.tech/topics/cms.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Automation](<https://devfeed.tech/topics/automation.md>), [WordPress](<https://devfeed.tech/topics/wordpress.md>)

Tags: [accessibility](<https://devfeed.tech/tags/accessibility.md>), [alt-text](<https://devfeed.tech/tags/alt-text.md>), [automation](<https://devfeed.tech/tags/automation.md>), [code](<https://devfeed.tech/tags/code.md>), [computer-vision](<https://devfeed.tech/tags/computer-vision.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [web-content-accessibility-guidelines](<https://devfeed.tech/tags/web-content-accessibility-guidelines.md>), [wordpress](<https://devfeed.tech/tags/wordpress.md>)

## AI overview

The article explains that effective alt text helps people using assistive technology understand images and is required as part of WCAG. Although machine learning and computer vision can automate image descriptions in content management systems, automated descriptions may be inaccurate or incomplete, so human review remains necessary.

## Source excerpt

Alternate (alt) text help people who use assistive technology understand images, and are a core part of the Web Content Accessibility Guidelines (WCAG). They require a human's input to be effective. To create alternate text, you add an alt attribute to an image that clearly and concisely describes what it is depicting: <img alt="A bowl of tomato soup and a grilled cheese sandwich." class="recipe-photo" src="/images/lunch.jpg" /> Assistive technology such as a screen reader will read the text provided inside the alt attribute's opening and closing quotes. It may sound something like: Image. A bowl of tomato soup and a grilled cheese sandwich. This allows a person to know an image is present, and what the image's content contains. With this context, someone using assistive technology can understand the image's purpose the same way someone who is not using assistive technology would. It's up to the person writing the code to add an alt attribute. It's also up to them to look at the image they're using and write an accurate description. Many Content Management Systems (CMS) allow the people who use them to include alt text by writing it in a dedicated input field. WordPress' alt text input field for uploaded images. Automation Technologies such as machine learning and computer vision allow us to guess at what an image is based on its arrangements of pixels. For example, we might show a computer hundreds of thousands of photos of fire hydrants to train it to understand what configurations of pixels look "fire hydrant-ish." The idea is that you can take these technologies and apply them to a CMS or other publishing system. The publishing system would ideally then be able to automatically apply alternate text to any image uploaded by an author. The problem is that automated image description is not a perfect science. Many of these automated detection technologies are still new and unsophisticated. This means there's a good chance they haven't been taught what the image you