Research map / 3D vision / 3D generation
Patch-Based 3D Natural Scene Generation From a Single Example
Weiyu Li, Xuelin Chen, Jue Wang, Baoquan Chen
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 |
|---|---|---|---|
| use patch-based exemplar synthesis with Plenoxels representation | Ours · St Alphage, Devil's Tower, Desert Lowpoly, Green Island, Stone Arch, Mountain, Vast Land | V-Qua.: St Alphage: 0.078 -> 0.022; Devil's Tower: 0.233 -> 0.032; Desert Lowpoly: 0.057 -> 0.020; Green Island: 0.254 -> 0.044; | Improved |
| replace TSDF geometry features with occupancy features (ablation) | Ours · St Alphage | V-Qua.: 0.022 -> 0.114 | Hurt |
| replace TSDF geometry features with occupancy features (ablation) | Ours · St Alphage | V-Div.: 0.312 -> 0.568 | Improved |
| replace coarse-to-fine exemplar training with recursive trilinear downsampling (ablation) | Ours · St Alphage | G-Qua.: 0.612 -> 0.847 | Hurt |
| use only value-based NNF synthesis (ablation) | Ours · St Alphage | V-Qua.: 0.022 -> 3.779 | Hurt |
| use only coordinate-based NNF synthesis (ablation) | Ours · St Alphage | G-Qua.: 0.612 -> 1.003 | Hurt |
| use only value-based NNF synthesis (ablation) | Ours · St Alphage | G-Div.: 0.473 -> 3.823 | Improved |
| use only coordinate-based NNF synthesis (ablation) | Ours · Devil's Tower | G-Qua.: 0.304 -> 0.256 | Improved |
| use only coordinate-based NNF synthesis (ablation) | Ours · Devil's Tower | G-Div.: 0.207 -> 0.492 | Hurt |
| use only approximate NNF instead of exact-to-approximate search (ablation) | Ours · 3D scene generation | visual quality | Hurt |
| use only exact NNF instead of exact-to-approximate search (ablation) | Ours · high-resolution 3D scene generation | computational feasibility | Hurt |
| increase the downscale ratio between pyramid levels (ablation) | Ours · 3D scene generation | visual quality | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- GRAF — GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis (NeurIPS 2020) · better on 4 of 4 shared results
- StyleNeRF — StyleNeRF: A Style-based 3D Aware Generator for High-resolution Image Synthesis (ICLR 2022) · 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
- Extend3D: Town-Scale 3D Generation (CVPR 2026 · similarity 0.89)
- 3D Scene Generation: A Survey (IJCV 2026 · similarity 0.88)
- A 3D Scene is Worth 1K Tokens: 3D-Grounded Representation for Scene Generation at Scale (NeurIPS 2026 · similarity 0.88)
- 3D-SceneDreamer: Text-Driven 3D-Consistent Scene Generation (CVPR 2024 · similarity 0.88)
- LT3SD: Latent Trees for 3D Scene Diffusion (CVPR 2025 · similarity 0.87)
- Real3D: The Curious Case of Neural Scene Degeneration (AAAI 2024 · similarity 0.87)
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.
