Research map / 3D vision / Gaussian splatting
Extrapolated Urban View Synthesis Benchmark
Xiangyu Han, Zhen Jia, Boyi Li, Yan Wang, Boris Ivanovic, Yurong You, Lingjie Liu, Yue Wang, Marco Pavone, Chen Feng, Yiming Li
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 |
|---|---|---|---|
| evaluate on extrapolated instead of interpolated views | 3DGS · EUVS Setting 1 | PSNR: 21.36 -> 16.37 | Hurt |
| evaluate on extrapolated instead of interpolated views | GSPro · EUVS Setting 1 | PSNR: 21.51 -> 16.39 | Hurt |
| evaluate on extrapolated instead of interpolated views | 3DGS · EUVS Setting 3 | PSNR: 21.22 -> 14.99 | Hurt |
| evaluate on extrapolated instead of interpolated views | GSPro · EUVS Setting 3 | PSNR: 21.58 -> 14.82 | Hurt |
| add diffusion priors (VEGS) to 3DGS | 3DGS · EUVS Setting 2 (rotation only) | PSNR: 19.53 -> 23.33 | Improved |
| add diffusion priors (VEGS) to 3DGS | 3DGS · EUVS Setting 2 (rotation only) | SSIM: 0.7511 -> 0.7949 | Improved |
| add diffusion priors (VEGS) to 3DGS | 3DGS · EUVS Setting 3 | PSNR: 14.99 -> 14.25 | No clear effect |
| add per-image appearance embedding (GS-W) | 3DGS-based baselines · EUVS Setting 3 | PSNR: 14.99 -> 19.36 | Mixed |
| add appearance embedding (GS-W) | Gaussian splatting · EUVS Setting 1 | PSNR: 16.37 -> 20.22 | Improved |
| increase number of training traversals (ablation) | GaussianPro · EUVS Setting 1 | PSNR/SSIM | Improved |
| use planar-based (PGSR/2DGS) instead of ellipsoid 3DGS | 3DGS · EUVS Setting 1 | PSNR: 16.37 -> 16.32 | No clear effect |
| use NeRF-based Zip-NeRF instead of 3DGS | 3DGS · EUVS Setting 2 | PSNR: 19.53 -> 17.36 | Hurt |
Closest papers by meaning
- Effective Multi-sensor Conditioning for Street-view Novel-view Synthesis (NeurIPS 2026 · similarity 0.88)
- Novel View Synthesis Under Large-Deviation Viewpoint for Autonomous Driving (AAAI 2025 · similarity 0.88)
- FreeVS: Generative View Synthesis on Free Driving Trajectory (ICLR 2025 · similarity 0.88)
- VR-Drive: Viewpoint-Robust End-to-End Driving with Feed-Forward 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.87)
- StreetCrafter: Street View Synthesis with Controllable Video Diffusion Models (CVPR 2025 · similarity 0.87)
- UMAMI: Unifying Masked Autoregressive Models and Deterministic Rendering for View Synthesis (NeurIPS 2025 · 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.
