Facebook envisions a future dominated by mobile video. So it's no surprise the company is working on a feature to automatically tag you and your friends to make those posts more searchable. Facebook’s Joaquin Quiñonero Candelae, the company's director of machine learning, went into detail about the feature today at the social network's F8 developer conference in San Francisco, according to TechCrunch.
Candelae says algorithms will soon be able to identify who's in the shot and tag the person appropriately so the moment is indexed forever on Facebook. That would allow you to go back in time and find any slice of any video in which a friend's face appeared on frame. Conceivably, you could also find whenever you showed up in someone else's video as well. It's likely Facebook is employing some of the same techniques it uses with DeepFace, its artificial intelligence-powered image recognition system.
Facebook is using AI to parse what's happening in user videos
The ultimate goal here is to reformat all of Facebook to make it more accommodating to video, both recorded and live-streamed. Facebook has been making an aggressive push into live video since late last year to compete with emerging platforms like Periscope and messaging apps like Snapchat. Facebook's vision is to allow anyone on Earth to share video through its app, whether with a 360-degree camera rig, a dedicated live-streaming device, or even a drone. Going even further — by understanding what's happening in those videos — the company can make its video platform even more powerful. However, it's unclear what type of commercial or advertising benefits this type of knowledge provides Facebook in the long run.