Back to results
US 12,608,883 B1Granted
Generating 3D models using neural networks
AssigneeNVIDIA Corporation
InventorsMing-Yu Liu, Chen-Hsuan Lin, Luming Tang, Tsung-Yi Lin
Life of this patent
Estimating the term…
Abstract
Apparatuses, systems, and techniques to generate three-dimensional models. In at least one embodiment, one or more neural networks are used to generate one or more three-dimensional models based, at least in part, on a single two-dimensional image.
Claims20
A processor comprising:
- one or more circuits to cause one or more neural networks to:
- generate, using a diffusion network of the one or more neural networks and based on a single two-dimensional (2D) image of a scene, a plurality of noisy images depicting a plurality of different simulated viewpoints of the scene;
- generate a plurality of denoised images from the plurality of noisy images;
- and
- generate a three-dimensional (3D) model of the scene based, at least in part, on the plurality of denoised images and 2D image of the scene.
2 more elements
A method comprising:
- causing one or more neural networks to:
- generate, using a diffusion network of the one or more neural networks and based on a single two-dimensional (2D) image of a scene, a plurality of noisy images depicting a plurality of different simulated viewpoints of the scene;
- generate a plurality of denoised images from the plurality of noisy images;
- and
- generate a three-dimensional (3D) model of the scene based, at least in part, on the plurality of denoised images and the single 2D image of the scene.
2 more elements
A system comprising:
- one or more processors to cause one or more neural networks to:
- generate, using a diffusion network of the one or more neural networks and based on a single two-dimensional (2D) image of a scene, a plurality of noisy images depicting a plurality of different simulated viewpoints of the scene;
- generate a plurality of denoised images from the plurality of noisy images;
- and
- generate a three-dimensional (3D) model of the scene based, at least in part, on the plurality of denoised images and the single 2D image of the scene.
2 more elements
Dependencies are inferred from the claim text.