Physically based shading in theory and practice

Hill, Stephen; McAuley, Stephen; Belcour, Laurent; Earl, Will; Harrysson, Niklas; Hillaire, Sébastien; Hoffman, Naty; Kerley, Lee; Patry, Jasmin; Pieké, Rob; Skliar, Igor; Stone, Jonathan; Barla, Pascal; Bati, Mégane; Georgiev, Iliyan

doi:10.1145/3388769.3407523

Cited by 4 publications

(2 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, h and w denote the spatial resolution of the textures, and c represents the total number of channels across all texture layers in the set. Increasing the visual fidelity of materials primarily relies on either adding more layers [HMB*20] or increasing their resolution which has led to a significant rise in memory requirements. To address this issue, textures are stored in GPU memory in a compressed format.…”

Section: Related Workmentioning

confidence: 99%

“…Besides that, micropolygon‐based data structures [MMT23, KSW21] allows the rendering of extremely detailed and dense scenes. In terms of visual aesthetics, the Physically Based Rendering (PBR) framework has been the standard in real‐time applications for well over a decade [MHH*12,MHM*13, HMB*20]. In this framework, a material is composed of multiple texture layers (such as albedo, normals, metalness, etc.)…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Real‐Time Neural Materials using Block‐Compressed Features

Weinreich,

De Oliveira,

Houdard

et al. 2024

Computer Graphics Forum

View full text Add to dashboard Cite

Neural materials typically consist of a collection of neural features along with a decoder network. The main challenge in integrating such models in real‐time rendering pipelines lies in the large size required to store their features in GPU memory and the complexity of evaluating the network efficiently. We present a neural material model whose features and decoder are specifically designed to be used in real‐time rendering pipelines. Our framework leverages hardware‐based block compression (BC) texture formats to store the learned features and trains the model to output the material information continuously in space and scale. To achieve this, we organize the features in a block‐based manner and emulate BC6 decompression during training, making it possible to export them as regular BC6 textures. This structure allows us to use high resolution features while maintaining a low memory footprint. Consequently, this enhances our model's overall capability, enabling the use of a lightweight and simple decoder architecture that can be evaluated directly in a shader. Furthermore, since the learned features can be decoded continuously, it allows for random uv sampling and smooth transition between scales without needing any subsequent filtering. As a result, our neural material has a small memory footprint, can be decoded extremely fast adding a minimal computational overhead to the rendering pipeline.

show abstract

Section: Related Workmentioning

confidence: 99%