Research map / 3D vision / NeRF and radiance fields
MatrixCity: A Large-scale City Dataset for City-scale Neural Rendering and Beyond
Yixuan Li, Lihan Jiang, Linning Xu, Yuanbo Xiangli, Zhenzhi Wang, Dahua Lin, Bo Dai
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 |
|---|---|---|---|
| naively train aerial and street data jointly (ablation) | MipNeRF-360 · MatrixCity fused data | PSNR: 23.05 -> 17.07 (street), aerial 28.37 | Hurt |
| naively train aerial and street data jointly (ablation) | TensoRF · MatrixCity fused data | PSNR: 22.10 -> 21.44 (street), aerial 27.26 | Hurt |
| increase street capture density from 5.0 m to 1.0 m (ablation) | Instant-NGP · MatrixCity street | PSNR: 21.436 -> 32.444 | Improved |
| increase street capture density from 5.0 m to 1.0 m (ablation) | MipNeRF-360 · MatrixCity street | PSNR: 27.75 -> 31.858 | Improved |
| model whole Small City aerial data with one model instead of five block models (ablation) | MipNeRF-360 · MatrixCity aerial | PSNR: 25.73 -> 24.41 | Hurt |
| model whole Small City aerial data with one model instead of five block models (ablation) | TensoRF · MatrixCity aerial | PSNR: 25.13 -> 25.09 | No clear effect |
| model whole Small City aerial data with one model instead of five block models (ablation) | Instant-NGP · MatrixCity aerial | PSNR: 25.69 -> 25.46 | No clear effect |
| split street data into smaller blocks (1 to 16 blocks) (ablation) | MipNeRF-360 · MatrixCity street Small City | PSNR: 22.00 -> 26.67 | Improved |
| increase model capacity from Small to Large at 16 blocks (ablation) | TensoRF · MatrixCity street Small City | PSNR: 22.96 -> 24.24 | Improved |
| increase model capacity from Small to Large at 16 blocks (ablation) | MipNeRF-360 · MatrixCity street Small City | PSNR: 24.64 -> 26.67 | Improved |
| use street data instead of aerial data for rendering | all baselines · MatrixCity | PSNR: aerial overall 25.73 vs street overall 23.24 (MipNeRF-360) | Hurt |
| use NeRF-W to filter dynamic objects in street scenes | NeRF-W · MatrixCity dynamic street | qualitative | Hurt |
Closest papers by meaning
- SUDS: Scalable Urban Dynamic Scenes (CVPR 2023 · similarity 0.92)
- Grid-Guided Neural Radiance Fields for Large Urban Scenes (CVPR 2023 · similarity 0.91)
- Multi-Level Neural Scene Graphs for Dynamic Urban Environments (CVPR 2024 · similarity 0.91)
- WildCity: A Real-World Dataset for City-Scale Rendering and Beyond (ECCV 2026 · similarity 0.90)
- Dynamic 3D Gaussian Fields for Urban Areas (NeurIPS 2024 · similarity 0.90)
- Swift-Mapping: Online Neural Implicit Dense Mapping in Urban Scenes (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.
