Research map / Video / Video generation
Comp-Attn: Present-and-Align Attention for Compositional Video Generation
Hongyu Zhang, Yufan Deng, Shenghai Yuan, Xuehan Hou, Yian Zhao, Peng Jin, Chang Liu, Jie Chen
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 Comp-Attn | Wan2.2-A14B · T2V-CompBench Avg | score: 0.5742 -> 0.6414 | Improved |
| add Comp-Attn | CogVideoX-5B · T2V-CompBench Avg | score: 0.4189 -> 0.5439 | Improved |
| add Comp-Attn | Wan2.1-14B · T2V-CompBench Avg | score: 0.5358 -> 0.6200 | Improved |
| add Comp-Attn | VideoCrafter2 · T2V-CompBench Avg | score: 0.4015 -> 0.4608 | Improved |
| use Comp-Attn instead of LVD | Wan2.2-A14B · T2V-CompBench Avg | score: 0.5988 -> 0.6414 | Improved |
| add SCI alone (ablation) | Wan2.2-A14B · T2V-CompBench Con-attr | score: 0.8394 -> 0.8613 | Improved |
| add LAM alone (ablation) | Wan2.2-A14B · T2V-CompBench Spatial | score: 0.6618 -> 0.7079 | Improved |
| add both SCI and LAM rather than SCI alone (ablation) | Wan2.2-A14B · T2V-CompBench Spatial | score: 0.6840 -> 0.7244 | Improved |
| replace fixed-rate interpolation with saliency-weighted SCI (ablation) | Wan2.2-A14B · T2V-CompBench Con-attr | score: 0.8426 -> 0.8613 | Improved |
| add IOU-guidance to LAM (ablation) | Wan2.2-A14B · T2V-CompBench | score | Improved |
| add Comp-Attn to Wan2.2 on VBench | Wan2.2-A14B · VBench Total | score: 0.8518 -> 0.8574 | Improved |
| add LVD to Wan2.2 on VBench | Wan2.2-A14B · VBench Total | score: 0.8518 -> 0.8360 | Hurt |
Closest papers by meaning
- VideoTetris: Towards Compositional Text-to-Video Generation (NeurIPS 2024 · similarity 0.87)
- Inference-Time Text-to-Video Alignment with Diffusion Latent Beam Search (NeurIPS 2025 · similarity 0.87)
- T2V-CompBench: A Comprehensive Benchmark for Compositional Text-to-video Generation (CVPR 2025 · similarity 0.87)
- R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image Generation (ICLR 2024 · similarity 0.86)
- VideoGrain: Modulating Space-Time Attention for Multi-Grained Video Editing (ICLR 2025 · similarity 0.86)
- VideoComp: Advancing Fine-Grained Compositional and Temporal Alignment in Video-Text Models (CVPR 2025 · 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.
