Generally, the images that we see uploaded on the various websites that we visit every day contain an alternative text (alt text) description necessary for screen readers, making those who suffer from some vision problem can benefit from the Text functions. to Voice. However, according to Microsoft, more than half of the images on the web are missing this alt text.
Screen readers can’t deal with an image that doesn’t have alt text, and Microsoft wanted to fix it with a new feature coming to Microsoft Edge. And it is that take advantage of machine learning to automatically generate this alt text and that screen readers have no trouble describing the image.
A great help for those who suffer from some kind of vision problem
Through an entry on the official Windows blog, Travis Leithead, Program Manager at Microsoft, Discuss the highlights of this feature coming to Edge.
“When a screen reader encounters an image without a label, that image can be automatically processed by machine learning algorithms to describe the image in words and capture any text it contains.”
Leithead explains that algorithms are not perfect, and that the quality of the descriptions will vary at timeshowever, for users using screen readers, it will sometimes be better than no context at all.
With this feature, Edge will send the images without a description to its Azure API for processing. The API will be able to create text in English, Spanish, Japanese, Portuguese, or Simplified Chinese, which will then be collected by screen readers. However, the browser will not be able to use this feature for images smaller than 50 x 50 pixels in resolution, very large images, or those that have been marked as decorative, pornographic, gore, or ‘sexually suggestive’.
This new feature Get right out of the box to Microsoft Edge for Windows, Mac, and Linux, leaving platforms such as Android or iOS behind for the time being. To access the function just type in the URL bar ‘edge://settings/accessibility’ and use Windows Narrator or another screen reader alternative to browse the web.