发表机构
Chalmers University of Technology; Zenseact; Linköping University（查尔姆斯理工大学; 森赛克特公司; 林雪平大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出VoroTracing，通过协同设计场景表示、优化与GPU执行，在RTX 5090上实现623 FPS的实时新视图合成，吞吐量优于现有光线方法和3D Gaussian Splatting，支持多种非针孔效应且保持竞争力。
AI 中文摘要
实时新视图合成目前主要由光栅化显式基元主导，这些基于投影的管线提供了高吞吐量，但对于畸变、卷帘快门、景深等非针孔效应需要专门的扩展。基于光线的渲染可自然地表达这些效应，但通常被认为速度太慢，无法实现具有竞争力的实时渲染。我们分析了可微Voronoi光线追踪中吞吐量的决定因素，确定遍历长度、每个单元的工作量和内存局部性是主要决定因素。基于此，我们引入VoroTracing，它协同设计场景表示、优化和GPU执行以降低这些成本：紧凑的八面体外观纹理减少内存流量，而集中于表面的不透明度促进早期终止；该固定预算表示无需剪枝或密集化即可优化，并通过专为相干遍历设计的GPU实现进行渲染。在Mip-NeRF 360数据集上，VoroTracing在RTX 5090上的渲染帧率达623 FPS，提供了比现有最快的基于光线的方法高3.2倍的吞吐量，比3D Gaussian Splatting高2.8倍，同时保持了有竞争力的重建质量。我们的渲染器通过光线生成和采样支持鱼眼、卷帘快门、运动模糊和景深效应，无需专门的光栅化。这些结果表明，基于光线的渲染可实现实时吞吐量，且具备灵活性。我们发布了源代码，详见此https URL。
英文摘要
Real-time novel view synthesis is dominated by rasterized explicit primitives. These projection-based pipelines provide high throughput but require specialized extensions for non-pinhole effects such as distortion, rolling shutter, and depth of field. Ray-based rendering expresses these effects naturally but is generally assumed too slow for competitive real-time rendering. We analyze the factors governing throughput in differentiable Voronoi ray tracing and identify traversal length, per-cell work, and memory locality as principal determinants. Guided by this, we introduce VoroTracing, which co-designs the scene representation, optimization, and GPU execution to reduce these costs. Compact octahedral appearance textures reduce memory traffic, while surface-concentrated opacity promotes early termination. The fixed-budget representation is optimized without pruning or densification and rendered with a GPU implementation designed for coherent traversal. On Mip-NeRF 360, VoroTracing renders at 623 FPS on an RTX 5090, providing $3.2\times$ the throughput of the fastest prior ray-based method and $2.8\times$ that of 3D Gaussian Splatting, while maintaining competitive reconstruction quality. Our renderer supports fisheye, rolling-shutter, motion-blur, and depth-of-field effects through ray generation and sampling, requiring no specialized rasterization. These results show that real-time throughput can be achieved with the flexibility of ray-based rendering. We release our source code, see https://research.zenseact.com/publications/vorotracing
