Paper

fVDB: A Deep-Learning Framework for Sparse, Large-Scale, and High-Performance Spatial Intelligence

Fuji-Tsang, Clement, Klár, Gergely, Ren, Xuanchi, Cong, Matthew, Museth, Ken, Swartz, Jonathan, Thakkar, Vijay, Sifakis, Eftychios, Williams, Francis, Huang, Jiahui, Li, Ruilong, Fidler, Sanja

2025

Overview

We present fVDB, a novel GPU-optimized framework for deep learning on large-scale 3D data. fVDB provides a complete set of differentiable primitives to build deep learning architectures for common tasks in 3D learning such as convolution, pooling, attention, ray-tracing, meshing, etc. fVDB simultaneously provides a much larger feature set (primitives and operators) than established frameworks with no loss in efficiency: our operators match or exceed the performance of other frameworks with narrower scope. Furthermore, fVDB can process datasets with much larger footprint and spatial resolution than prior works, while providing a competitive memory footprint on small inputs. To achieve this combination of versatility and performance, fVDB relies on a single novel VDB index grid acceleration structure paired with several key innovations including GPU accelerated sparse grid construction, convolution using tensorcores, fast ray tracing kernels using a Hierarchical Digital Differential Analyzer algorithm (HDDA), and jagged tensors. Our framework is fully integrated with PyTorch enabling interoperability with existing pipelines, and we demonstrate its effectiveness on a number of representative tasks such as large-scale point-cloud segmentation, high resolution 3D generative modeling, unbounded scale Neural Radiance Fields, and large-scale point cloud reconstruction.

• Publisher: Cornell University Library, arXiv.org
• Subject: Algorithms / Analog computers / Convolution / Deep learning / Differential analyzers / Effectiveness / Image reconstruction / Image segmentation / Operators (mathematics) / Ray tracing / Spatial resolution / Tensors / Three dimensional models