Research map / Video / Video generation
Aligning Global Semantics and Local Textures in Generative Video Enhancement
Zhikai Chen, Fuchen Long, Zhaofan Qiu, Ting Yao, Wengang Zhou, Jiebo Luo, Tao Mei
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 image reference via vanilla cross-attention (ablation) | GenVE basic model A · YouHQ40 | DOVER: 77.29 -> 78.75 | Improved |
| add image reference via vanilla cross-attention (ablation) | GenVE basic model A · YouHQ40 | MUSIQ: 68.34 -> 69.10 | Improved |
| replace vanilla cross-attention with local-aware cross-attention (LCA) (ablation) | GenVE model B · YouHQ40 | DOVER: 78.75 -> 79.98 | Improved |
| replace vanilla cross-attention with LCA (ablation) | GenVE model B · YouHQ40 | MUSIQ: 69.10 -> 69.45 | Improved |
| add noise augmentation (ablation) | GenVE model C · YouHQ40 | DOVER: 79.98 -> 80.16 | Improved |
| add noise augmentation (ablation) | GenVE model C · YouHQ40 | MUSIQ: 69.45 -> 69.53 | Improved |
| add masking condition (ablation) | GenVE model C1 · YouHQ40 | DOVER: 80.16 -> 82.15 | Improved |
| add masking condition (ablation) | GenVE model C1 · YouHQ40 | MUSIQ: 69.53 -> 69.88 | Improved |
| add temporal augmentation (ablation) | GenVE model C2 · YouHQ40 | DOVER: 82.15 -> 83.50 | Improved |
| add temporal augmentation (ablation) | GenVE model C2 · YouHQ40 | MUSIQ: 69.88 -> 69.93 | Improved |
| add semantics-aligned HQ image reference to LQ-video-only guidance | GenVE vs VEnhancer · YouHQ40 | DOVER: 77.08 -> 83.50 | Improved |
| add semantics-aligned HQ image reference to LQ-video-only guidance | GenVE vs VEnhancer · YouHQ40 | MUSIQ: 68.99 -> 69.93 | Improved |
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.
- Upscale-A-Video — Upscale-A-Video: Temporal-Consistent Diffusion Model for Real-World Video Super-Resolution (CVPR 2024) · better on 3 of 3 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
- Generative Video Propagation (CVPR 2025 · similarity 0.87)
- FreeLong: Training-Free Long Video Generation with SpectralBlend Temporal Attention (NeurIPS 2024 · similarity 0.87)
- GenRec: Unifying Video Generation and Recognition with Diffusion Models (NeurIPS 2024 · similarity 0.86)
- GenFusion: Closing the Loop between Reconstruction and Generation via Videos (CVPR 2025 · similarity 0.86)
- CustomCrafter: Customized Video Generation with Preserving Motion and Concept Composition Abilities (AAAI 2025 · similarity 0.86)
- Upscale-A-Video: Temporal-Consistent Diffusion Model for Real-World Video Super-Resolution (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.
