AI Describes Photos on Facebook for Users with Vision Impairments

Facebook can now automatically create alternative text for images generating descriptions that enable users who are blind or have low vision to envisage the content of the photo. The iOS app provides an audio breakdown of what’s happening in the photo using object recognition technology.

Using its vast supply of user images, Facebook has trained a deep neural network driving a computer vision system to recognize object in images. As is a standard in the WCAG 2.0 guidelines, the results are translated to “alt text,” which can be read by screen readers. 

Share This Post