Research map / Video / Video generation
Video Generation From Single Semantic Label Map
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use the two-stage seg2vid approach instead of FG | single-label-map video generation on Cityscapes · Cityscapes validation | FID: 3.69 -> 3.52 | Helped |
| Use the two-stage seg2vid approach instead of MoCoGAN | single-label-map video generation on Cityscapes · Cityscapes validation | FID: 8.77 -> 3.52 | Helped |
| Use the two-stage seg2vid approach instead of vid2vid | single-label-map video generation on Cityscapes · Cityscapes validation | FID: 4.86 -> 3.52 | Helped |
| Use seg2vid instead of FG | video prediction from a single starting frame · Cityscapes validation | FID: 2.86 -> 1.80 | Helped |
| Use seg2vid instead of MoCoGAN | video prediction from a single starting frame · Cityscapes validation | FID: 7.06 -> 1.80 | Helped |
| Add optical-flow-based motion representation (ablation) | seg2vid · Cityscapes video generation | visual quality | Helped |
| Add semantic label maps to flow prediction (ablation) | seg2vid · Cityscapes video generation | foreground and background motion | Helped |
| Separate semantic maps into foreground and background groups (ablation) | seg2vid with semantic label maps · Cityscapes video generation | motion structure detail | Helped |
| Use seg2vid instead of vid2vid | single-label-map video generation on Cityscapes · Cityscapes user study | human preference score: 0.37 vs 0.63 | Hurt |
Closest papers by meaning
- Decouple Content and Motion for Conditional Image-to-Video Generation (AAAI 2024 · similarity 0.85)
- StoryDiffusion: Consistent Self-Attention for Long-Range Image and Video Generation (NeurIPS 2024 · similarity 0.85)
- SEINE: Short-to-Long Video Diffusion Model for Generative Transition and Prediction (ICLR 2024 · similarity 0.85)
- Taming generative video models for zero-shot optical flow extraction (NeurIPS 2025 · similarity 0.85)
- FreeLong: Training-Free Long Video Generation with SpectralBlend Temporal Attention (NeurIPS 2024 · similarity 0.84)
- EIDT-V: Exploiting Intersections in Diffusion Trajectories for Model-Agnostic, Zero-Shot, Training-Free Text-to-Video Generation (CVPR 2025 · 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.