Up until now Facebook 3D photo feature was only available for only those had higher end Android and iOS smartphones with dual-lens “portrait mode”. Those who had single, rear facing camera capabilities couldn’t enjoy this feature.
Therefore, Facebook felt a need to train a model that could infer 3D positions of a wide variety of photos. Furthermore, it planned to develop a system that could be used on typical mobile processors in a fraction of a second.
Now with the help of Facebook machine learning 3D photos can be virtually inferred from any 2D picture immaterial of whether the shot has been taken from a standard single lens camera. The same feature can also be used on images taken several years back and have been uploaded on laptop or smartphone.
All About The Facebook 3D Photo Feature:
How Has Facebook Made This 2D to 3D Conversion Of Image Possible
To make a Facebook photo 3D, engineers at Facebook are using convolutional neural nets which are also known as CNN. It is because of CNN that this 3D technology is accessible to all those with single-lens tablets or smartphones.
Another question that might strike your mind is how does CNN make 2D to 3D conversion of image possible on Facebook?
For the purpose of training the CNN large amounts of data is gathered from public 3D photos. Convolutional Neural Network or CNN estimates the distance of each pixel. Then, an architectural network is built using a set of mobile-optimized neural building blocks. An automated architecture finds an apt configuration of these blocks because of which the system performs the task in under a second. To further minimize any quality degradation, quantization process is carried out.
Note: The information has been gathered from Facebook AI Blog
What About Those Who Have High-End Smartphones
For those who have smartphones that have high-end cameras, the Facebook 3D photo technology could prove to be even better as now they can take 3D selfies with the help of their front-facing cameras. More specifically anyone with a midrange or higher Android device or iOS 7 can enjoy Facebook 3D photo feature
How Can Users Use 3D Photo Feature?
- You can start utilizing this feature by first updating your app and then looking for the ‘3D photo’ option which appears when you create a new post.
- Tap on the 3D photo option and select an image. Now, Facebook will generate a 3D image
It must be noted that the image that is generated is a mere simulation since it doesn’t contain the full depth map of the photo that has been captured with the help of your dual-lens camera. You will be able to see the effect when you tilt the smartphone.
Is Facebook Photo Pioneer In This?
Much before Facebook enabled users to infer 3D images from 2D images, Google has been doing it for years with Pixel. Also, there are many apps that do the same job. But something that makes the Facebook 3D photo feature special is that Facebook is getting this feature to more people and that too for free. You won’t have to spend oodles to convert photos from 2D to 3D
So, Ready, Set, Convert?
By converting 2D photos to 3D you will now be able to see your images pop up and come to life. Isn’t that great? With that being said, do let us know if you liked the post and for more fun filled tech related content keep reading Tweak Library. You can also find us on Facebook and YouTube as well.