18132714. HIGH-FIDELITY THREE-DIMENSIONAL ASSET ENCODING simplified abstract (Adobe Inc.)

HIGH-FIDELITY THREE-DIMENSIONAL ASSET ENCODING

Organization Name

Adobe Inc.

Inventor(s)

Krishna Bhargava Mullia Lakshminarayana of San Francisco CA (US)

Valentin Deschaintre of London (GB)

Nathan Carr of San Jose CA (US)

Milos Hasan of San Jose CA (US)

Bailey Miller of Pittsburgh PA (US)

HIGH-FIDELITY THREE-DIMENSIONAL ASSET ENCODING - A simplified explanation of the abstract

This abstract first appeared for US patent application 18132714 titled 'HIGH-FIDELITY THREE-DIMENSIONAL ASSET ENCODING

The patent application focuses on training a neural material to render images and applying the material map to a coarse geometry to encode high-fidelity assets efficiently.

• Training involves sampling lighting and camera configurations to render an image of a target asset.
• Optimization of a loss function compares the target asset with the neural material to train it for high-fidelity encoding.
• The neural material is restricted to a predetermined geometry for reproducible assets suitable for mobile devices or web deployment with limited computational resources.

1. 1. 1. Potential Applications:

This technology can be used in various industries such as gaming, virtual reality, augmented reality, and e-commerce for realistic image rendering.

1. 1. 1. Problems Solved:

The technology addresses the challenge of efficiently encoding high-fidelity assets for deployment on devices with limited computational capabilities.

1. 1. 1. Benefits:

The benefits include the production of highly detailed images on resource-constrained devices, enhancing user experience and visual quality.

1. 1. 1. Commercial Applications:

"Neural Material Rendering for High-Fidelity Asset Encoding" can revolutionize the gaming industry, virtual reality applications, e-commerce product visualization, and more by providing realistic and detailed images efficiently.

1. 1. 1. Prior Art:

Researchers can explore prior art related to neural material rendering, image encoding, and asset optimization techniques in computer graphics and machine learning fields.

1. 1. 1. Frequently Updated Research:

Stay updated on advancements in neural material rendering, image encoding algorithms, and asset optimization methods to leverage the latest technologies for high-fidelity image production.

1. 1. 1. 1. Questions about Neural Material Rendering for High-Fidelity Asset Encoding:

1. How does neural material training improve image rendering quality?

  - Neural material training enhances image rendering quality by optimizing the loss function to encode high-fidelity assets efficiently.

2. What are the potential applications of this technology beyond gaming and virtual reality?

  - This technology can also be applied in e-commerce for realistic product visualization and in architectural design for detailed rendering of structures.

Original Abstract Submitted

Certain aspects and features of this disclosure relate to rendering images by training a neural material and applying the material map to a coarse geometry to provide high-fidelity asset encoding. For example, training can involve sampling for a set of lighting and camera configurations arranged to render an image of a target asset. A value for a loss function comparing the target asset with the neural material can be optimized to train the neural material to encode a high-fidelity model of the target asset. This technique restricts the application of the neural material to a specific predetermined geometry, resulting in a reproducible asset that can be used efficiently. Such an asset can be deployed, as examples, to mobile devices or to the web, where the computational budget is limited, and nevertheless produce highly detailed images.