Facebook's 3D photos expand to millions more phones thanks to AI
Figuring out 3D scenes from 2D imagery can pep up your posts -- and the technology can help Tesla cars and robots navigate the real world.
Stephen Shanklandprincipal writer
Stephen Shankland has been a reporter at CNET since 1998 and writes about processors, digital photography, AI, quantum computing, computer science, materials science, supercomputers, drones, browsers, 3D printing, USB, and new computing technology in general. He has a soft spot in his heart for standards groups and I/O interfaces. His first big scoop was about radioactive cat poop.
Expertiseprocessors, semiconductors, web browsers, quantum computing, supercomputers, AI, 3D printing, drones, computer science, physics, programming, materials science, USB, UWB, Android, digital photography, scienceCredentials
I've been covering the technology industry for 24 years and was a science writer for five years before that. I've got deep expertise in microprocessors, digital photography, computer hardware and software, internet standards, web technology, and other dee
has used artificial intelligence technology to bring 3D photos to people taking pictures on phones that only can take plain old 2D shots. By training its AI systems on existing 3D photos, Facebook can figure out how to make 3D photos even if your camera can't.
Facebook announced 3D photos in 2018 but initially offered the technology only with
that could take portrait mode shots. For those,
uses dual cameras to reconstruct 3D data about the scene, called a depth map. Since then it expanded to other dual-camera phones like newer
models. But now Facebook creates that depth map on its own using the AI technology.
"This advance makes 3D photo technology easily accessible for the first time to the many millions of people who use single-lens camera phones or
," Facebook researchers said in a Friday blog post explaining the 3D technology. Another benefit: people can make 3D photos with single front-facing cameras, too, Facebook said.
It won't work on every smartphone, but it should work on an
or newer phone, or a recent midrange or better Android device, Facebook said.
Making 3D out of 2D photos is a nice feature for Facebook and companies like
, which uses AI to figure out depth so it can blur backgrounds in portrait photos. But 3D depth maps have more serious uses, too. Tesla uses AI-powered depth judgments to power its Autopilot and Full Self-Driving features without using expensive 3D laser scanner systems called lidar. And you don't want your car to misjudge the distance to anything around it when you're cruising down the highway at 70 mph.
For its part, Facebook plans to improve the 3D technology with new abilities, like calculating the geometric orientation of surfaces and using the technology in augmented reality. It also hopes others can use Facebook's approach for challenges like building
that navigate the real world.