Research map / 3D vision / 3D generation
Tiled Multiplane Images for Practical 3D Photography
Numair Khan, Lei Xiao, Douglas Lanman
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 |
|---|---|---|---|
| use confidence-weighted k-means instead of linear plane spacing (ablation) | TMPI plane placement · Spaces depth discretization | MAE (x1e-3): 48.7 -> 27.1 | Improved |
| use confidence-weighted k-means instead of vanilla k-means (ablation) | TMPI plane placement · Spaces depth discretization | MAE (x1e-3): 35.3 -> 27.1 | Improved |
| use confidence-weighted k-means instead of linear plane spacing (ablation) | TMPI plane placement · Tanks and Temples depth discretization | MAE (x1e-3): 46.0 -> 27.0 | Improved |
| use confidence-weighted k-means instead of vanilla k-means (ablation) | TMPI plane placement · Tanks and Temples depth discretization | MAE (x1e-3): 35.7 -> 27.0 | Improved |
| use confidence-weighted k-means instead of vanilla k-means (ablation) | TMPI plane placement · Spaces depth discretization | MSE (x1e-3): 2.00 -> 1.20 | Improved |
| use confidence-weighted k-means instead of linear spacing (ablation) | TMPI plane placement · Spaces depth discretization | MSE (x1e-3): 3.80 -> 1.20 | Improved |
| use smaller tiles and more planes (up to 8) (ablation) | TMPI plane placement · view synthesis quality | quality | Improved |
| increase planes beyond 8 (ablation) | TMPI plane placement · view synthesis quality | quality | No clear effect |
| use 4 adaptive planes per tile instead of 32 planes per image (TMPI vs AdaMPI) | TMPI · Tanks and Temples | PSNR: 18.62 -> 18.69 | No clear effect |
| use 4 adaptive planes per tile instead of 32 planes per image (TMPI vs AdaMPI) | TMPI · Spaces | PSNR: 26.17 -> 24.93 | Hurt |
Closest papers by meaning
- Structural Multiplane Image: Bridging Neural View Synthesis and 3D Reconstruction (CVPR 2023 · similarity 0.90)
- Pushing the Boundaries of View Extrapolation With Multiplane Images (CVPR 2019 · similarity 0.89)
- SAMPLING: Scene-adaptive Hierarchical Multiplane Images Representation for Novel View Synthesis from a Single Image (ICCV 2023 · similarity 0.87)
- DeepView: View Synthesis With Learned Gradient Descent (CVPR 2019 · similarity 0.86)
- Multiplane Prior Guided Few-Shot Aerial Scene Rendering (CVPR 2024 · similarity 0.83)
- Stereo Vision Conversion from Planar Videos Based on Temporal Multiplane Images (AAAI 2024 · similarity 0.83)
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.
