3D Scene Rendering With Guide-Channel Denoising at Mixed Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Path-tracing in 3-D graphics is computationally expensive and inefficient for reducing noise in rendered images, especially on devices with limited resources, as increasing the number of samples required for noise reduction is impractical.
Innovation Solution
A method involving machine learning models to derive enhanced guide channels from initial guide channels, using neural networks to approximate noisy images as a function of these enhanced channels, allowing for efficient denoising and rendering at higher framerates and quality, particularly on mobile devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of samples is increased to reduce noise in path-traced images, then the quality of the rendered image is improved, but the computational cost increases significantly
Solution Approach 1:
The patent introduces guide channels as intermediary data structures that capture scene geometry and lighting information. These guide channels serve as mediators between the noisy path-traced image and the denoised output, enabling noise reduction without requiring additional path-tracing samples. The guide channels are rendered once and reused during the denoising process, significantly reducing computational cost compared to increasing sample counts.
Solution Approach 2:
The patent replaces the mechanical approach of reducing noise by increasing sample counts with a computational approach using machine learning models. Instead of mechanically tracing more light paths, the system uses neural networks to predict and remove noise based on guide channel information, substituting a more efficient computational method for the traditionally resource-intensive sampling approach.
2Productivity
If guide channels are used for denoising, then computational efficiency is improved, but the availability of suitable guide channels with accurate scene structure is required
Solution Approach 1:
The patent makes guide channels universal by rendering them for all pixels in the scene simultaneously using rasterization techniques. Rather than computing guide channel information individually for each pixel (which would be computationally expensive), the system renders comprehensive guide channels that cover the entire scene once, making them universally available for denoising any region of the image without additional computational overhead.
3Measurement precision
If denoising is performed at full resolution, then image quality is maintained, but memory requirements and computational load increase
Solution Approach 1:
The patent segments the denoising process into two distinct stages: guide channel rendering at full resolution, and denoising computation at reduced resolution. By separating these operations, the system maintains full-resolution guide channels for accuracy while performing the computationally intensive denoising calculations at lower resolution, significantly reducing memory requirements and computational load during the denoising phase.
Data Source
AI summary
An image of a 3-D scene is rendered by rendering a noisy image at a first resolution; obtaining initial guide channels at the first resolution, and obtaining corresponding initial guide channels at a second resolution. When the two resolutions are the same, the initial guide channels at the first resolution and the corresponding initial guide channels at the second resolution may be provided by a single set of initial guide channels. Enhanced guide channels are derived from the initial guide channels using machine learning models. For each of a plurality of local neighbourhoods, the parameters of a denoising model that approximates the noisy image (in the local neighbourhood) are calculated as a function of the enhanced guide channels (at the first resolution), and the calculated parameters are applied to the one or more enhanced guide channels (at the second resolution), to produce a denoised image at the second resolution.