A voice from Facebook's new technology identifies a photo on a user's timeline and reads the name of the person who posted a photo, the date and time the photo was uploaded, the photo's caption and the amount of likes and comments the photo has received aloud. The voice also identifies items that may be in the object using advanced technology.

For example, the tool may say, "Image may contain: tree, sky, outdoor" or "One or more people. Jewelry. Smiling. Nineteen likes. Three comments."

"Now I can see the picture in my head, like, 'Yeah you shouldn't have been that close up,'" a blind Facebook user said.

Facebook is calling this new feature automatic alternative text. Before this, blind users would only hear the name of the person who posted the image and the word "photo" when they scrolled past images in their news feed.

The company's object recognition technology reportedly holds billions of parameters, which allows the new automatic alternative text tool to say what's in each photo.

Right now, the alternative text is only available in English and accessible on iOS devices, but Facebook said it plans to add other languages and platforms soon.

"Even having three words just helps flesh out all the details that I can't see. That makes me feel included, and like I'm a part of it too," another blind Facebook user said.