Learning Generic Local Shape Properties for Adaptive Super-Sampling

C. Reinbold¹, R. Westermann¹

¹Computer Graphics & Visualization Group, Technische Universität München, Germany

Overview of PatchNet. Given a scene represented by a sparse voxel octree, its local geometry patches are encoded in low-dimensional feature vectors called patch codes. Given a ray-voxel intersection p, its surrounding patch codes are interpolated to derive an abstract descriptor of the local geometry surrounding p. Together with the view-dependent rendering parameters θ, the subsequent decoder interprets the interpolated patch code and derives occupancy and luminance.

Abstract

We propose a novel encoder/decoder-based neural network architecture that learns view-dependent shape and appearance of geometry represented by voxel representations. Since the network is trained on local geometry patches, it generalizes to arbitrary models. A geometry model is first encoded into a sparse voxel octree of features learned by a network, and this model representation can then be decoded by another network in-turn for the intended task. We utilize the network for adaptive supersampling in ray-tracing, to predict super-sampling patterns when seeing coarse-scale geometry. We discuss and evaluate the proposed network design, and demonstrate that the decoder network is compact and can be integrated seamlessly into on-chip ray-tracing kernels. We compare the results to previous screen-space super-sampling strategies as well as non-network-based world-space approaches.

Associated publications

Learning Generic Local Shape Properties for Adaptive Super-Sampling

C. Reinbold, R. Westermann

Eurographics 2022 [PDF] [BIB]