Research map / 3D vision / NeRF and radiance fields
Temporal-MPI: Enabling Multi-Plane Images for Dynamic Scene Modelling via Temporal Basis Learning
Wenpeng Xing, Jie Chen
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 low-frequency component (ablation) | Temporal-MPI · Nvidia Dynamic Scenes | PSNR: 10.3 -> 24.87 | Improved |
| add high-frequency components and temporal basis (ablation) | Temporal-MPI · Nvidia Dynamic Scenes | PSNR: 22.6 -> 24.87 | Improved |
| add low-frequency component (ablation) | Temporal-MPI · Nvidia Dynamic Scenes | SSIM: 0.192 -> 0.859 | Improved |
| add high-frequency components (ablation) | Temporal-MPI · Nvidia Dynamic Scenes | LPIPS: 0.213 -> 0.196 | Improved |
| increase number of timestamps from 8 to 60 (ablation) | Temporal-MPI · Skating-2 | PSNR: 30.323 -> 25.012 | Hurt |
| increase number of timestamps from 8 to 60 (ablation) | Temporal-MPI · Jumping | PSNR: 25.850 -> 23.854 | Hurt |
| increase number of timestamps from 8 to 60 (ablation) | Temporal-MPI · Balloon2-2 | PSNR: 25.775 -> 23.945 | Hurt |
| increase basis from 5 to 13 at 596x320 (ablation) | Temporal-MPI · inference speed | inference time (s): 0.002 -> 0.003 | No clear effect |
| increase planes from 32 to 192 at 1038x1940 (ablation) | Temporal-MPI · inference speed | inference time (s): 0.025 -> 0.030 | No clear effect |
| use shared temporal basis instead of per-frame MPIs (ablation) | Temporal-MPI · Dynamic Scenes storage (T=24) | storage (Mb): 5400 -> 481 | Improved |
| use shared temporal basis instead of per-frame LLFF MPIs | Temporal-MPI · Dynamic Scenes | PSNR: 35.41 -> 31.94 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- D-NeRF — D-NeRF: Neural Radiance Fields for Dynamic Scenes (CVPR 2021) · better on 2 of 2 shared results
- SynSin — SynSin: End-to-End View Synthesis From a Single Image (CVPR 2020) · better on 2 of 2 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
- Fast View Synthesis of Casual Videos with Soup-of-Planes (ECCV 2024 · similarity 0.87)
- Neural Scene Chronology (CVPR 2023 · similarity 0.86)
- View Synthesis of Dynamic Scenes Based on Deep 3D Mask Volume (TPAMI 2023 · similarity 0.85)
- NeX: Real-Time View Synthesis With Neural Basis Expansion (CVPR 2021 · similarity 0.85)
- Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting (ICLR 2024 · similarity 0.85)
- DaReNeRF: Direction-aware Representation for Dynamic Scenes (CVPR 2024 · similarity 0.84)
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.
