Plenoxels: Radiance Fields without Neural Networks
About
We introduce Plenoxels (plenoptic voxels), a system for photorealistic view synthesis. Plenoxels represent a scene as a sparse 3D grid with spherical harmonics. This representation can be optimized from calibrated images via gradient methods and regularization without any neural components. On standard, benchmark tasks, Plenoxels are optimized two orders of magnitude faster than Neural Radiance Fields with no loss in visual quality.
Alex Yu, Sara Fridovich-Keil, Matthew Tancik, Qinhong Chen, Benjamin Recht, Angjoo Kanazawa• 2021
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Novel View Synthesis | Tanks&Temples (test) | PSNR27.43 | 239 | |
| Novel View Synthesis | Mip-NeRF 360 (test) | PSNR23.62 | 166 | |
| Novel View Synthesis | LLFF | PSNR26.29 | 124 | |
| Novel View Synthesis | Mip-NeRF360 | PSNR23.08 | 104 | |
| Novel View Synthesis | Mip-NeRF 360 | PSNR23.08 | 102 | |
| Novel View Synthesis | NeRF Synthetic | PSNR31.76 | 92 | |
| Novel View Synthesis | LLFF (test) | PSNR26.29 | 79 | |
| Novel View Synthesis | Deep Blending (test) | PSNR23.06 | 64 | |
| Novel View Synthesis | ScanNet | PSNR22.35 | 58 | |
| Novel View Synthesis | Mip-NeRF360 (test) | PSNR23.08 | 58 |
Showing 10 of 41 rows