LogIn"Our model leverages the permutation equivariance of the transformer when conditioning on the partial scene, and is trained to be permutation-invariant across object orderings. Our model is trained end-to-end as an autoregressive generative model using only labeled 3D bounding boxes as supervision," wrote the team. "Our formulation allows applying a single trained model to automatic layout synthesis and to a number of interactive scenarios with versatile user input, such as automatic placement of user-provided objects, object suggestion with user-provided constraints, and room completion."
What does the whole thing mean in a few words? Well, you could potentially train such a system using a set of examples, give it an IKEA catalog and make it generate new interiors in a matter of seconds to decorate new spaces. That might sound a bit scary for designers, yes, but these systems still need an initial style.
The official page features the team's paper on the matter, a number of examples, and more. You can learn more about the project and find the code here.
Don't forget to join our new Reddit page, our new Telegram channel, follow us on Instagram and Twitter, where we are sharing breakdowns, the latest news, awesome artworks, and more.