Radiance Meshes for Volumetric Reconstruction
About
We introduce radiance meshes, a technique for representing radiance fields with constant density tetrahedral cells produced with a Delaunay tetrahedralization. Unlike a Voronoi diagram, a Delaunay tetrahedralization yields simple triangles that are natively supported by existing hardware. As such, our model is able to perform exact and fast volume rendering using both rasterization and ray-tracing. We introduce a new rasterization method that achieves faster rendering speeds than all prior radiance field representations (assuming an equivalent number of primitives and resolution) across a variety of platforms. Optimizing the positions of Delaunay vertices introduces topological discontinuities (edge flips). To solve this, we use a Zip-NeRF-style backbone which allows us to express a smoothly varying field even when the topology changes. Our rendering method exactly evaluates the volume rendering equation and enables high quality, real-time view synthesis on standard consumer hardware. Our tetrahedral meshes also lend themselves to a variety of exciting applications including fisheye lens distortion, physics-based simulation, editing, and mesh extraction.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Novel View Synthesis | Tanks&Temples | PSNR23.13 | 95 | |
| Novel View Synthesis | Deep Blending | PSNR29.39 | 48 | |
| 3D Scene Reconstruction | MipNeRF360 Indoor (test) | PSNR29.29 | 15 | |
| Novel View Synthesis | MipNeRF360 | PSNR27.15 | 13 | |
| 3D Scene Reconstruction | MipNeRF360 Outdoor (test) | -- | 8 |