Lay-A-Scene: Personalized 3D Object Arrangement Using Text-to-Image Priors

TL;DR

Lay-A-Scene introduces a novel method leveraging pre-trained text-to-image models to generate and infer plausible 3D object arrangements in scenes with multiple high-resolution objects, addressing open-set challenges.

Contribution

It presents a personalized approach to 3D scene arrangement using text-to-image priors, enabling the placement of unseen objects without neglecting any in the scene.

Findings

Often generates coherent 3D arrangements

Achieves feasible object placements in complex scenes

Validated with Objaverse objects and human raters

Abstract

Generating 3D visual scenes is at the forefront of visual generative AI, but current 3D generation techniques struggle with generating scenes with multiple high-resolution objects. Here we introduce Lay-A-Scene, which solves the task of Open-set 3D Object Arrangement, effectively arranging unseen objects. Given a set of 3D objects, the task is to find a plausible arrangement of these objects in a scene. We address this task by leveraging pre-trained text-to-image models. We personalize the model and explain how to generate images of a scene that contains multiple predefined objects without neglecting any of them. Then, we describe how to infer the 3D poses and arrangement of objects from a 2D generated image by finding a consistent projection of objects onto the 2D scene. We evaluate the quality of Lay-A-Scene using 3D objects from Objaverse and human raters and find that it often generates coherent and feasible 3D object arrangements.