Research map / Video / Video segmentation and tracking
Video Frame Interpolation With Many-to-Many Splatting and Spatial Selective Refinement
Ping Hu, Simon Niklaus, Lu Zhang, Stan Sclaroff, Kate Saenko
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 motion refinement network (ablation) | M2M with PWC-Net · Vimeo90K | PSNR: 33.97 -> 34.94 | Improved |
| add motion refinement network (ablation) | M2M with DISFlow · Vimeo90K | PSNR: 31.93 -> 34.32 | Improved |
| add joint flow encoding (JFE) (ablation) | M2M with PWC-Net MRN · Vimeo90K | PSNR: 34.94 -> 35.09 | Improved |
| add low-rank feature modulation (LFM) (ablation) | M2M with PWC-Net MRN · Vimeo90K | PSNR: 34.94 -> 35.07 | Improved |
| add both JFE and LFM (ablation) | M2M with PWC-Net · Vimeo90K | PSNR: 34.94 -> 35.15 | Improved |
| add reliability score in fusion (ablation) | M2M with PWC-Net · Vimeo90K | PSNR: 35.15 -> 35.24 | Improved |
| add reliability score in fusion (ablation) | M2M with DISFlow · Vimeo90K | PSNR: 34.78 -> 34.93 | Improved |
| increase sub-motion vectors from N=1 to N=4 (ablation) | M2M with PWC-Net · Vimeo90K | PSNR: 35.24 -> 35.40 | Improved |
| increase sub-motion vectors from N=4 to N=8 (ablation) | M2M with PWC-Net · Vimeo90K | PSNR: 35.40 -> 35.39 | No clear effect |
| add spatial selective refinement (100% patches) | PWC-M2M · Vimeo90K | PSNR: 35.40 -> 36.19 | Improved |
| add SSR instead of enlarging M2M 2.5x (ablation) | M2M++ 100% · Vimeo90K | PSNR: 35.81 -> 36.19 | Improved |
| enlarge M2M channels 2x (ablation) | PWC-M2M · Vimeo90K | PSNR: 35.40 -> 35.72 | Improved |
Closest papers by meaning
- Many-to-Many Splatting for Efficient Video Frame Interpolation (CVPR 2022 · similarity 0.98)
- EDEN: Enhanced Diffusion for High-quality Large-motion Video Frame Interpolation (CVPR 2025 · similarity 0.88)
- Disentangled Motion Modeling for Video Frame Interpolation (AAAI 2025 · similarity 0.87)
- HiCoM: Hierarchical Coherent Motion for Dynamic Streamable Scenes with 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.87)
- ViBiDSampler: Enhancing Video Interpolation Using Bidirectional Diffusion Sampler (ICLR 2025 · similarity 0.87)
- Sparse Global Matching for Video Frame Interpolation with Large Motion (CVPR 2024 · similarity 0.87)
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.
