3inGAN: Learning a 3D Generative Model from Images of a Self-similar Scene
This addresses the challenge of generating diverse 3D scenes from limited data for applications in graphics and vision, representing a novel but incremental advance in 3D GANs.
The paper tackles the problem of learning a 3D generative model from 2D images of a single self-similar scene, enabling the production of view-consistent 3D scene variations without flickering or artifacts, and demonstrates feasibility with qualitative and quantitative comparisons against recent methods.
We introduce 3inGAN, an unconditional 3D generative model trained from 2D images of a single self-similar 3D scene. Such a model can be used to produce 3D "remixes" of a given scene, by mapping spatial latent codes into a 3D volumetric representation, which can subsequently be rendered from arbitrary views using physically based volume rendering. By construction, the generated scenes remain view-consistent across arbitrary camera configurations, without any flickering or spatio-temporal artifacts. During training, we employ a combination of 2D, obtained through differentiable volume tracing, and 3D Generative Adversarial Network (GAN) losses, across multiple scales, enforcing realism on both its 3D structure and the 2D renderings. We show results on semi-stochastic scenes of varying scale and complexity, obtained from real and synthetic sources. We demonstrate, for the first time, the feasibility of learning plausible view-consistent 3D scene variations from a single exemplar scene and provide qualitative and quantitative comparisons against recent related methods.