Research map / Video / Video segmentation and tracking
Asymmetric Bilateral Motion Estimation for Video Frame Interpolation
Junheum Park, Chul Lee, Chang-Su Kim
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 reliability mask to ABMR-Net (ablation) | ABMR-Net · Vimeo90K (warped frames) | PSNR: 27.81 -> 28.32 | Improved |
| add offset map to ABMR-Net (ablation) | ABMR-Net · Vimeo90K (warped frames) | PSNR: 27.81 -> 28.40 | Improved |
| add both reliability mask and offset map to ABMR-Net (ablation) | ABMR-Net · Vimeo90K (warped frames) | PSNR: 27.81 -> 28.80 | Improved |
| add both reliability mask and offset map to ABMR-Net (ablation) | ABMR-Net · SNU-FILM Extreme (warped frames) | PSNR: 22.75 -> 23.07 | Improved |
| add asymmetric candidate warped frames to symmetric ones (ablation) | ABME synthesis · Vimeo90K | PSNR: 35.91 -> 36.18 | Improved |
| add symmetric candidates to asymmetric ones (ablation) | ABME synthesis · Vimeo90K | PSNR: 36.05 -> 36.18 | Improved |
| add asymmetric candidates to symmetric ones (ablation) | ABME synthesis · SNU-FILM Extreme | PSNR: 25.28 -> 25.42 | Improved |
| add filtered feature map as RefineNet input (ablation) | RefineNet · Vimeo90K | PSNR: reduced by 0.1 dB when removed | Improved |
| use asymmetric bilateral motion field instead of symmetric (SBMF) (ablation) | warped-frame motion fields · Vimeo90K, Easy, Medium | PSNR: more than 2 dB | Improved |
| use asymmetric bilateral motion field instead of symmetric (SBMF) (ablation) | warped-frame motion fields · SNU-FILM Hard and Extreme | PSNR: at least 1 dB | Improved |
| ABME versus BMBC | video frame interpolation · Vimeo90K | PSNR: 35.01 -> 36.18 | Improved |
| ABME versus BMBC | video frame interpolation · SNU-FILM Easy | PSNR: 39.90 -> 39.59 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- BMBC — BMBC: Bilateral Motion Estimation with Bilateral Cost Volume for Video Interpolation (ECCV 2020) · better on 5 of 6 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
- BMBC: Bilateral Motion Estimation with Bilateral Cost Volume for Video Interpolation (ECCV 2020 · similarity 0.90)
- BiFormer: Learning Bilateral Motion Estimation via Bilateral Transformer for 4K Video Frame Interpolation (CVPR 2023 · similarity 0.89)
- Video Frame Interpolation without Temporal Priors (NeurIPS 2020 · similarity 0.87)
- AMT: All-Pairs Multi-Field Transforms for Efficient Frame Interpolation (CVPR 2023 · similarity 0.86)
- ViBiDSampler: Enhancing Video Interpolation Using Bidirectional Diffusion Sampler (ICLR 2025 · similarity 0.85)
- Perception-Oriented Video Frame Interpolation via Asymmetric Blending (CVPR 2024 · similarity 0.85)
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.
