Research map / 3D vision / Gaussian splatting
Neural Harmonic Textures for High-Quality Primitive Based Neural Reconstruction
Jorge Condor, Nicolas Moënne-Loccoz, Merlin Nimier-David, Piotr Didyk, Zan Gojcic, Qi Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace spherical harmonics appearance with Neural Harmonic Textures | 3DGUT-MCMC · MipNeRF360 (standardized, 1M primitives) | PSNR: 27.93 -> 28.46 | Improved |
| replace Spherical Voronoi appearance with Neural Harmonic Textures | 3DGUT-MCMC · Tanks & Temples (standardized) | PSNR: 24.18 -> 24.79 | Improved |
| add NHT to 2DGS | 2DGS · MipNeRF360 | PSNR: 27.48 -> 28.27 | Improved |
| add NHT to Triangle Splatting | Triangle Splatting · MipNeRF360 | SSIM: 0.808 -> 0.807 | No clear effect |
| add NHT to Triangle Splatting | Triangle Splatting · MipNeRF360 | PSNR: 27.00 -> 27.52 | Improved |
| add NHT to 3DGUT-MCMC | 3DGUT-MCMC · MipNeRF360 | LPIPS: 0.247 -> 0.232 | Improved |
| add NHT to 3DGUT-MCMC | 3DGUT-MCMC · MipNeRF360 (standardized) | FPS: 201 -> 140 | Hurt |
| add NHT to 3DGUT-MCMC | 3DGUT-MCMC · MipNeRF360 training time | minutes: 7.1 -> 14.5 | Hurt |
| increase per-primitive features from 48 to 64 (ablation) | 3DGUT-MCMC + NHT · MipNeRF360 | PSNR: 28.65 -> 28.68 | Improved |
| increase per-primitive feature dimension from 4 to 64 (ablation) | NHT semantic reconstruction · Tanks & Temples truck, RGB+LSEG | PSNR: 23.81 -> 26.42 | Improved |
| train and evaluate on JPEG references instead of downscaled PNGs (ablation) | NHT · MipNeRF360 | PSNR: -0.284 | Hurt |
| use NHT instead of explicit BBSplat texture at equal primitive count | BBSplat · MipNeRF360 300K primitives | PSNR: 27.09 -> 27.83 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ZipNeRF — Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields (ICCV 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Learning Efficient Fuse-and-Refine for Feed-Forward 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.84)
- HyRF: Hybrid Radiance Fields for Memory-efficient and High-quality Novel View Synthesis (NeurIPS 2025 · similarity 0.84)
- Structure Consistent Gaussian Splatting with Matching Prior for Few-shot Novel View Synthesis (NeurIPS 2024 · similarity 0.84)
- Learning Spatially Collaged Fourier Bases for Implicit Neural Representation (AAAI 2024 · similarity 0.84)
- Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting (ICLR 2024 · similarity 0.84)
- CompactSplat: Spatially Adaptive Gaussian Distribution for Feedforward Scene Reconstruction (NeurIPS 2026 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
