Facebook has upgraded its 3D Photos feature to let you create a 3D image out of any shot. Previously, the feature only worked with the portrait mode images taken from devices with two or more cameras, but the social network has now expanded the
tool to include support for images taken from single-camera phones, as well.
tool to include support for images taken from single-camera phones, as well.
How 3D Photos used to work before -
When the feature, used to create and share 3D photos, debuted in October 2018, it was largely dependent on the portrait mode shots taken by two or more cameras of a device.
Basically, the portrait photos were used by Facebook's AI to stitch different layers together and make a typical 2D shot three-dimensional, complete with the "depth" and movement of a 3D photo.
Facebook is using machine learning for single-camera phones
Now, to upgrade the tool for single-camera photos, Facebook has deployed a machine learning-based technique.
The system uses machine learning to infer the three-dimensional structure of an image and add depth and 3D effect according to it.
This ultimately transforms the image into a 3D post, without even using a portrait mode shot from a second camera.
This would even work with selfies, says Facebook
Facebook says that the upgraded 3D Photos tool should work with images taken via iPhone 7 or higher or "a recent midrange or better Android device."
It might take a while to roll out on all phones, but once that happens, it will be relatively easier to share 3D posts.
Plus, you would also be able to share depth-rich 3D selfies on your Facebook.
How to use the feature?
In order to use the feature, you will have to update the Facebook app and follow the regular process to post content on the News Feed.
Just navigate to "What's on your mind", select "3D Photos", and pick any image of your choice - portrait or not - to post it with a 3D effect, maybe even with a caption.
Images won't be truly three-dimensional
It's also worth noting that the feature will not make images truly three-dimensional. You won't be able to rotate a subject to see it from all angles, but there will be enough depth to scroll, pan, tilt, and see it from a few different perspectives.
