Facebook’s AI to Help Blind People ‘See’
Facebook on Monday started using artificial intelligence to help people with illustration enjoy photos posted at the social network. It introduced machine learning technology taught to recognize objects in pictures and then explain photos clearly. With AI, this app automatically produces an explanation of each photo a user comes across. When they’re using a screen reader on iOS, they’ll be able to listen to a list of items in the pictures, such as the image may a explanation like “Image may contain three people, smiling, outdoors.”
“As Facebook becomes an increasingly visual experience, we hope our new automatic alternative text technology will help the blind community experience Facebook the same way others enjoy it,” said accessibility specialist Matt King.
“While this technology is still nascent, tapping its current capabilities to describe photos is a huge step toward providing our visually impaired community the same benefits and enjoyment that everyone else gets from photos,” Matt King said.
The descriptions, called Automatic Alternative Text, are generated as image alt text. It’s a standard HTML attribute designed for enlarged convenience that screen reading tools can easily pick up. The AI is behind this uses Facebook’s object recognition technology to identify what’s in a photo. The company said that it took 10 months to develop the feature to its current stage, using neural networks that mimic the brain to learn how to recognize images after being skilled with millions of examples.
Automatic alternative text is now available in English for iOS users in the US, UK, Canada, Australia, and New Zealand. Facebook says it will soon come out to more platforms, languages, and countries.