Research map / Point clouds / Point cloud classification
View Synthesis with Sculpted Neural Points
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 SNP instead of NeRF | View synthesis · DTU | LPIPS: 0.266 -> 0.156 | Helped |
| use SNP instead of NeRF | View synthesis · DTU | PSNR: 28.97 -> 26.68 | Hurt |
| use SNP instead of NeRF | View synthesis · LLFF | LPIPS: 0.250 -> 0.229 | Helped |
| use SNP instead of Point-NeRF | View synthesis · NeRF-Synthetic (all 8 scenes) | PSNR: 33.31 -> 27.47 | Hurt |
| use SNP instead of NeRF | View synthesis · NeRF-Synthetic | Inference (FPS): 0.053 -> 5.06 | Helped |
| replace proposed pruning with DCC filtering (ablation) | SNP · DTU | PSNR: 26.68 -> 19.97 | Hurt |
| add point pruning and point adding (ablation) | SNP without adding or pruning · DTU | PSNR: 25.06 -> 26.68 | Helped |
| add point adding (ablation) | SNP with pruning only · DTU | PSNR: 26.15 -> 26.68 | Helped |
| allow gradient-based refinement of point positions and opacity (ablation) | SNP with fixed point positions and opacity · DTU | PSNR: 26.52 -> 26.68 | Helped |
| add spherical-harmonic view dependence (ablation) | SNP without view dependence · DTU | PSNR: 25.67 -> 26.68 | Helped |
| replace MLP view dependence with spherical harmonics (ablation) | SNP with MLP view dependence · DTU | PSNR: 26.30 -> 26.68 | Helped |
| add point dropout (ablation) | SNP without point dropout · DTU | PSNR: 25.40 -> 26.68 | Helped |
Closest papers by meaning
- SynSin: End-to-End View Synthesis From a Single Image (CVPR 2020 · similarity 0.88)
- NPBG++: Accelerating Neural Point-Based Graphics (CVPR 2022 · similarity 0.88)
- HashPoint: Accelerated Point Searching and Sampling for Neural Rendering (CVPR 2024 · similarity 0.88)
- PDF: Point Diffusion Implicit Function for Large-scale Scene Neural Representation (NeurIPS 2023 · similarity 0.88)
- Boosting View Synthesis With Residual Transfer (CVPR 2022 · similarity 0.87)
- Neural Point Light Fields (CVPR 2022 · 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.