Research map / Video / Video segmentation and tracking
SVI360: Spherical Video Interpolation
Le-Kim NGUYEN, Renato Martins, Pascal Vasseur, Cedric Demonceaux
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 spherical weighted Charbonnier loss (ablation) | SVI360 · FlowScape | PSNR: 33.99 -> 38.68 | Improved |
| add census loss (ablation) | SVI360 · FlowScape | PSNR: 38.34 -> 38.68 | Improved |
| add spherical multi-field refinement with orthogonal candidates (ablation) | SVI360 · FlowScape | PSNR: 38.08 -> 38.68 | Improved |
| add Spherical Refiner block (ablation) | SVI360 · FlowScape | PSNR: 38.59 -> 38.68 | Improved |
| add latitude weighting to Charbonnier loss (ablation) | SVI360 · FlowScape | WS-PSNR: 37.63 -> 37.73 | Improved |
| add latitude weighting to Charbonnier loss (ablation) | SVI360 · FlowScape | PSNR: 38.72 -> 38.68 | Hurt |
| add spherical weighted loss (ablation) | AMT-G · FlowScape | PSNR: 38.05 -> 38.29 | Improved |
| use smaller AMT backbones (LiteSVI360, 22.9M params) (ablation) | SVI360 · FlowScape | PSNR: 38.68 -> 38.55 | No clear effect |
| replace AMT-G with dual-branch SVI360 | AMT-G · FlowScape | PSNR: 38.05 -> 38.68 | Improved |
| replace AMT-G with SVI360 | AMT-G · 360VFI Extreme | WS-PSNR: 25.55 -> 25.92 | Improved |
| replace AMT-G with SVI360 | AMT-G · Flow360 | PSNR: 33.77 -> 34.02 | Improved |
| replace AMT-G with SVI360 | AMT-G · 360VFI Easy | WS-PSNR: 34.84 -> 34.81 | No clear effect |
Closest papers by meaning
- Video Frame Interpolation without Temporal Priors (NeurIPS 2020 · similarity 0.86)
- Disentangled Motion Modeling for Video Frame Interpolation (AAAI 2025 · similarity 0.85)
- Imagine360: Immersive 360 Video Generation from Perspective Anchor (NeurIPS 2025 · similarity 0.85)
- LAVIB: A Large-scale Video Interpolation Benchmark (NeurIPS 2024 · similarity 0.84)
- Perception-Oriented Video Frame Interpolation via Asymmetric Blending (CVPR 2024 · similarity 0.84)
- OCAI: Improving Optical Flow Estimation by Occlusion and Consistency Aware Interpolation (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.
