ESVR: 3D Ellipsoid-based Sparse Volume Rendering via Structure-aware Primitive Learning and Per-primitive Ray Sampling
Authors
Suemin Jeon (Korea University), Youjin Kim (Korea University), Jungwoo Park (Korea University), Kyungryun Lee (Korea University), Won-Ki Jeong (Korea University)
Presentation
- Session
- Turn up the volume(s)!
- Time
- Thursday, Nov 12, 13:12 – 13:24 (US/Eastern) · session 13:00 – 14:30
- Location
- Hall America north
Links
Sign in to access the preprint PDF.
Sign in- Download Supplemental Material
Keywords
3D Gaussian Splatting, Scientific Visualization, Volume Visualization
Abstract
Efficient representation and rendering of large-scale sparse volumetric data remain challenging in scientific visualization, as meaningful structures often occupy only a small fraction of the spatial domain. While direct volume rendering (DVR) provides high-quality visualization, its computational and memory costs scale poorly with data size. Recent advances in 3D Gaussian Splatting (3DGS) address this challenge by representing volumetric scenes with compact geometric primitives, enabling efficient, high-fidelity rendering. However, existing 3DGS-based methods learn from DVR rendered images rather than raw volumes, leading to information loss and limiting flexible transfer function control for interactive exploration. To address these limitations, we propose ESVR, an ellipsoid-based sparse volume rendering framework that directly learns and renders volumetric data in 3D space. Our method combines differentiable ellipsoidal primitives with bounded support, structure-aware primitive learning with complementary pruning, and a per-primitive ray sampling strategy for fast and accurate transfer function mapping. To support large-scale datasets, we further introduce a chunk-based optimization scheme with ghost ellipsoids, providing boundary context during training. Across large sparse datasets, ESVR achieves up to four orders of magnitude compression and real-time rendering at 43–223 FPS while maintaining competitive reconstruction quality.
For Practitioners
This paper may interest practitioners in scientific visualization, large-scale scientific computing, medical and biological imaging, and GPU-accelerated rendering. These practitioners often need to store, render, and interactively explore volumetric datasets in which important structures occupy only a small portion of an extremely large spatial domain. From this work, they can learn how adaptive primitive-based representations can reduce storage and rendering costs while preserving fine structures, supporting flexible transfer-function exploration and real-time visualization of large sparse volumes.