Research map / Video / Video generation
World-Consistent Video-to-Video Synthesis
Arun Mallya, Ting-Chun Wang, Karan Sapra, Ming-Yu Liu
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 guidance images (world consistency) via Multi-SPADE (ablation) | Ours w/o W.C. · MannequinChallenge | FID: 27.23 -> 22.69 | Improved |
| add guidance images (world consistency) (ablation) | Ours w/o W.C. · Cityscapes | FID: 51.51 -> 49.89 | Hurt |
| add guidance images (world consistency) (ablation) | Ours w/o W.C. · Cityscapes | mIOU: 0.62 -> 0.61 | Hurt |
| add guidance images (world consistency) (ablation) | Ours w/o W.C. · ScanNet | FID: 20.93 -> 21.07 | Hurt |
| add guidance images (world consistency) (ablation) | Ours w/o W.C. · ScanNet | mIOU: 0.12 -> 0.13 | Improved |
| add guidance images (world consistency) (ablation) | Ours w/o W.C. · MannequinChallenge | mIOU: 0.17 -> 0.19 | Improved |
| add guidance images (world consistency) (ablation) | Ours w/o W.C. · ScanNet short-term consistency | warping L1 error: 0.0278 -> 0.0192 | Improved |
| add guidance images (world consistency) (ablation) | Ours w/o W.C. · Cityscapes short-term consistency | warping L1 error: 0.0032 -> 0.0029 | Improved |
| replace vid2vid with the proposed method | vid2vid · Cityscapes | FID: 69.07 -> 49.89 | Improved |
| replace vid2vid with the proposed method | vid2vid · MannequinChallenge | FID: 72.25 -> 22.69 | Improved |
| replace vid2vid with the proposed method | vid2vid · MannequinChallenge forward-backward consistency | delta RGB: 37.56 -> 12.61 | Improved |
| replace vid2vid with the proposed method | vid2vid · ScanNet forward-backward consistency | delta RGB: 46.30 -> 11.85 | Improved |
Closest papers by meaning
- Geometry-guided Online 3D Video Synthesis with Multi-View Temporal Consistency (CVPR 2025 · similarity 0.87)
- FlowVid: Taming Imperfect Optical Flows for Consistent Video-to-Video Synthesis (CVPR 2024 · similarity 0.87)
- SimVS: Simulating World Inconsistencies for Robust View Synthesis (CVPR 2025 · similarity 0.87)
- WorldWeaver: Generating Long-Horizon Video Worlds via Rich Perception (NeurIPS 2025 · similarity 0.86)
- SpatialDreamer: Self-supervised Stereo Video Synthesis from Monocular Input (CVPR 2025 · similarity 0.86)
- MultiDiff: Consistent Novel View Synthesis from a Single Image (CVPR 2024 · similarity 0.86)
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.
