Research map / 3D vision / NeRF and radiance fields
CompNVS: Novel View Synthesis with Scene Completion
Zuoyue Li, Tianxing Fan, Zhenqiang Li, Zhaopeng Cui, Yoichi Sato, Marc Pollefeys, Martin R. Oswald
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 |
|---|---|---|---|
| add point cloud encoder (ablation) | basic framework (B) · Replica | SSIM: 0.730 -> 0.746 | Improved |
| add point cloud encoder (ablation) | basic framework (B) · Replica | LPIPS: 0.418 -> 0.395 | Improved |
| add point cloud encoder (ablation) | basic framework (B) · Replica | FID: 128.5 -> 135.0 | Hurt |
| add disentangled alpha/RGB MLPs (ablation) | B+E · Replica | FID: 135.0 -> 124.2 | Improved |
| add disentangled alpha/RGB MLPs (ablation) | B+E · Replica | SSIM: 0.746 -> 0.729 | Hurt |
| add disentangled alpha/RGB MLPs (ablation) | B+E · Replica | LPIPS (unobserved): 0.628 -> 0.672 | Hurt |
| add 2D refinement module (ablation) | B+E+D · Replica | FID: 124.2 -> 74.84 | Improved |
| add 2D refinement module (ablation) | B+E+D · Replica | LPIPS: 0.404 -> 0.332 | Improved |
| add 2D refinement module (ablation) | B+E+D · Replica | SSIM: 0.729 -> 0.706 | Hurt |
| add 2D refinement module (ablation) | B+E+D · Replica | FID (unobserved): 216.3 -> 182.6 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SPSG — SPSG: Self-Supervised Photometric Scene Generation From RGB-D Scans (CVPR 2021) · better on 6 of 6 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
- SG-NN: Sparse Generative Neural Networks for Self-Supervised Scene Completion of RGB-D Scans (CVPR 2020 · similarity 0.91)
- Novel View Synthesis from A Few Glimpses via Test-Time Natural Video Completion (NeurIPS 2025 · similarity 0.90)
- GenRec: Knowing Where to Reconstruct and Where to Generate (NeurIPS 2026 · similarity 0.89)
- Novel View Synthesis with Pixel-Space Diffusion Models (CVPR 2025 · similarity 0.88)
- UMAMI: Unifying Masked Autoregressive Models and Deterministic Rendering for View Synthesis (NeurIPS 2025 · similarity 0.88)
- Decompositional Neural Scene Reconstruction with Generative Diffusion Prior (CVPR 2025 · similarity 0.88)
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.
