Research map / Point clouds / Point cloud classification
PAPR: Proximity Attention Point Rendering
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 PAPR instead of NeRF | novel-view rendering · NeRF Synthetic | PSNR: 31.00 -> 32.07 | Helped |
| Use PAPR instead of Gaussian Splatting | point-based novel-view rendering · NeRF Synthetic | LPIPS_vgg: 0.084 -> 0.038 | Helped |
| Use PAPR instead of Gaussian Splatting | point-based novel-view rendering · Tanks & Temples | PSNR: 26.81 -> 28.72 | Helped |
| Add point position p_i to key features (ablation) | PAPR ray-dependent point embedding · Lego | PSNR: 31.67 -> 32.62 | Helped |
| Add point position p_i to key features (ablation) | PAPR ray-dependent point embedding · Mic | PSNR: 32.00 -> 35.64 | Helped |
| Add point position p_i and displacement t_i,j to embedding (ablation) | PAPR ray-dependent point embedding · Lego | PSNR: 13.26 -> 32.62 | Helped |
| Add point position p_i and displacement t_i,j to embedding (ablation) | PAPR ray-dependent point embedding · Mic | PSNR: 18.73 -> 35.64 | Helped |
| Use more points (ablation) | PAPR · NeRF Synthetic | LPIPS | Helped |
| Add point pruning (ablation) | PAPR point cloud optimization · Lego | point cloud outliers | Helped |
| Use ray-relative attention contributions normalized across points instead of absolute-distance splat contributions | learning point positions from scratch · Lego | geometry reconstruction | Helped |
Closest papers by meaning
- PAPR in Motion: Seamless Point-level 3D Scene Interpolation (CVPR 2024 · similarity 0.93)
- Differentiable Point-based Inverse Rendering (CVPR 2024 · similarity 0.85)
- UniPre3D: Unified Pre-training of 3D Point Cloud Models with Cross-Modal Gaussian Splatting (CVPR 2025 · similarity 0.85)
- Point2Real: Bridging the Gap between Point Cloud and Realistic Image for Open-World 3D Recognition (AAAI 2024 · similarity 0.85)
- On Geometry-Enhanced Parameter-Efficient Fine-Tuning for 3D Scene Segmentation (NeurIPS 2025 · similarity 0.85)
- HashPoint: Accelerated Point Searching and Sampling for Neural Rendering (CVPR 2024 · similarity 0.85)
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.