Synthesizing realistic and diverse indoor 3D scene layouts in a controllable fashion opens up applications in simulated navigation and virtual reality. As concise and robust representations of a scene, scene graphs have proven to be well-suited as the semantic control on the generated layout. We present a variant of the conditional variational autoencoder (cVAE) model to synthesize 3D scenes from scene graphs and floor plans. We exploit the properties of self-attention layers to capture high-level relationships between objects in a scene, and use these as the building blocks of our model. Our model, leverages graph transformers to estimate the size, dimension and orientation of the objects in a room while satisfying relationships in the given scene graph. Our experiments shows self-attention layers leads to sparser (7.9x compared to Graphto3D) and more diverse scenes (16%).
@article{arxiv.2404.01887,
title = {3D scene generation from scene graphs and self-attention},
author = {Pietro Bonazzi and Mengqi Wang and Diego Martin Arroyo and Fabian Manhardt and Nico Messikomer and Federico Tombari and Davide Scaramuzza},
journal= {arXiv preprint arXiv:2404.01887},
year = {2024}
}
Comments
Some authors were not timely informed of the submission