Research map / 3D vision / 3D generation
Learning To Render Novel Views From Wide-Baseline Stereo Pairs
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 the full method rather than GPNR | Novel view synthesis from wide-baseline stereo pairs · RealEstate10K test | PSNR: 18.55 -> 21.38 | Helped |
| Use the full method rather than IBRNet | Novel view synthesis from wide-baseline stereo pairs · ACID test | PSNR: 19.24 -> 23.63 | Helped |
| Replace uniform 3D sampling with 2D epipolar-line sampling (ablation) | Base Model · RealEstate10K test | PSNR: 18.11 -> 19.02 | Helped |
| Add cross-image correspondence matching (ablation) | 2D Sampling variant · RealEstate10K test | PSNR: 19.02 -> 19.52 | Helped |
| Replace monocular encoding with a multiview encoder (ablation) | Cross Correspondence variant · RealEstate10K test | PSNR: 19.52 -> 20.43 | Helped |
| Add regularization loss for multiview consistency (ablation) | Multiview Encoder variant · RealEstate10K test | PSNR: 20.43 -> 19.84 | Hurt |
| Add regularization loss for multiview consistency (ablation) | Multiview Encoder variant · RealEstate10K test | SSIM: 0.794 -> 0.808 | Helped |
| Add geometrically consistent data augmentation (ablation) | Regularization Loss variant · RealEstate10K test | PSNR: 19.84 -> 21.38 | Helped |
| Increase volumetric samples from 64 to 192 instead of using epipolar samples (ablation) | Volumetric renderer · RealEstate10K test | PSNR: 20.60 vs. 21.38 | Hurt |
| Increase the number of context views from 2 to 3 (ablation) | Ours · RealEstate10K test | PSNR: 21.38 -> 22.29 | Helped |
| Increase the baseline from 32 to 128 (ablation) | Ours · RealEstate10K test | PSNR: 26.24 -> 21.38 | Hurt |
Closest papers by meaning
- LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias (ICLR 2025 · similarity 0.90)
- Novel View Synthesis from A Few Glimpses via Test-Time Natural Video Completion (NeurIPS 2025 · similarity 0.90)
- Zero-Shot Novel View and Depth Synthesis with Multi-View Geometric Diffusion (CVPR 2025 · similarity 0.90)
- Novel View Synthesis with Pixel-Space Diffusion Models (CVPR 2025 · similarity 0.90)
- UMAMI: Unifying Masked Autoregressive Models and Deterministic Rendering for View Synthesis (NeurIPS 2025 · similarity 0.90)
- Sparse3D: Distilling Multiview-Consistent Diffusion for Object Reconstruction from Sparse Views (AAAI 2024 · similarity 0.90)
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.