Google Has Developed a New 3D Object Recognition Process, Which Could Lead to Improved AR Experiences

As recently demonstrated by Snapchat, AR systems are advancing, and facilitating new options for branding and advertising which utilize more immersive, engaging visual elements.

But one of the limitations of AR is 3D modeling, and the capacity for smartphone cameras to accurately scale object size in order to match them with virtual comparisons, and provide realistic depictions that can be layered over real-world scenes.

That’s where this new process from Google comes in – this week on the Google Engineering Blog, the company has outlined its latest advance in 3D modeling, which could see a significant shift in AR rendering in the near future.

As explained by Google:

Today, we are announcing the release of MediaPipe Objectron, a mobile real-time 3D object detection pipeline for everyday objects. This pipeline detects objects in 2D images, and estimates their poses and sizes through a machine learning (ML) model, trained on a newly created 3D dataset.”

The process improves the capacity to capture an object’s size, position and orientation in the world, which, as noted, could have significant impacts on the accuracy of AR applications, leading to more impressive, useful digital placement tools.

“Our novel approach, called AR Synthetic Data Generation, places virtual objects into scenes that have AR session data, which allows us to leverage camera poses, detected planar surfaces, and estimated lighting to generate placements that are physically probable and with lighting that matches the scene. This approach results in high-quality synthetic data with rendered objects that respect the scene geometry and fit seamlessly into real backgrounds.”

Google MediaPipe example

We’ve already seen examples of AR product placement in real-world environments, including various make-up try-on tools and IKEA’s AR furniture display app.

Those initial usages are impressive, but more advanced 3D modeling capacity like this will enable even smarter virtual placement, and better, more accurate tools. That could be major for eCommerce and virtual shopping, with more ways to see what objects look like in your world before you make a purchase, while there are also advanced potential opportunities for brand activations and engagement through improved AR tools.

Google is sharing its latest advances with the developer community, with a view to future product enhancements. And with reports also suggesting that Apple is also testing a new, dedicated AR app, and potentially its own AR glasses, we could be in for some major AR updates in the latter half of 2020.  

You can read more about Google’s MediaPipe process here.

Like this article?

Share on facebook
Share on Facebook
Share on twitter
Share on Twitter
Share on linkedin
Share on Linkdin
Share on pinterest
Share on Pinterest

Leave a comment