The technology aims to speed up naming people in the 100 million photos uploaded to Facebook each day.
Facebook Thursday began testing face-detection technology for its popular photo-tagging feature.
In the coming months, Facebook plans to enhance all areas of users' interaction with photographs, said Sam Odio, product manager for photos, in a company blog. After all, 99% of Facebook users have uploaded at least one photograph, said Odio, who joined Facebook two months ago after the social networking company acquired his photo startup Divvyshot.
"More than 100 million photos are uploaded every day. That's insane," he wrote. "This might surprise you, but within the photos product most people spend their time uploading, browsing, and tagging photos. As a result, we're working to improve your experience in each of those areas. For tagging, we began testing today a faster and easier way for you to tag your own photos."
Under the existing process, users upload photos, then click on each subject's face and tag that individual, continuing until all the people are named and the entire album is tagged. When working with group shots or large albums, the process can be unwieldy and time-consuming. Using Facebook's face detection enabled solution, a person's face pops up and a box asks the user to enter the friend's name.
"With this new feature, tagging is faster since you don't need to select a face. It's already selected for you, just like those rectangles you see around your friends' faces when you take a photo with a modern digital camera. All that's left for you to do is type a name and hit enter," said Odio.
Currently, Facebook is running face-detection as a limited trial so not all Facebook users will have access to this feature. Facebook did not provide a timetable for a wider roll-out of its face-detection capability.
"The tagging feature is just the start of improvements we're trying out," Odio said.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?