Research map / Video / Video generation
Stylized Video Generation via Decoupled Data Synthesis and Gated Style Token Injection
Xin Wei, Yijie Fang, Yanjia Li, Liangyi Wu, Qin Yang, Mingrui Zhu, Nannan Wang, Xinbo Gao
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 |
|---|---|---|---|
| use clean style latent at timestep 0 instead of noisy (ablation) | IGST · Style-guided T2V ablation | CSD: 0.1490 -> 0.5219 | Improved |
| add head-wise gated attention (ablation) | IGST · Style-guided T2V ablation | CSD: 0.4530 -> 0.5219 | Improved |
| disable RoPE for style tokens (ablation) | IGST · Style-guided T2V ablation | CLIP text score: 0.2866 -> 0.3010 | Improved |
| add Diffusion-DPO post-training (ablation) | IGST · Style-guided T2V ablation | CSD: 0.4910 -> 0.5219 | Improved |
| add Diffusion-DPO post-training (ablation) | IGST · Style-guided T2V ablation | Dynamic Degree: 0.7792 -> 0.8247 | Improved |
| use token concatenation instead of encoder-based explicit injection (ablation) | IGST · Style-guided T2V ablation | CSD: 0.3969 -> 0.5219 | Improved |
| train StyleMaster-style explicit injection on MoStyle-5K videos (ablation) | StyleMaster · Style-guided T2V ablation | Dynamic Degree: 0.1710 -> 0.6472 | Improved |
| add image-video mixed training instead of video-only (ablation) | IGST · Style-guided T2V ablation | qualitative appearance | Improved |
| use IGST instead of StyleMaster | IGST · Style-guided T2V ablation | Image Quality: 0.6735 -> 0.6568 | Hurt |
| use IGST instead of StyleMaster | IGST · Style-guided T2V ablation | Dynamic Degree: 0.1710 -> 0.8247 | Improved |
| use clean style latent (CLIP text score) (ablation) | IGST · Style-guided T2V ablation | CLIP text score: 0.3144 -> 0.3010 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- StyleMaster — StyleMaster: Stylize Your Video with Artistic Generation and Translation (CVPR 2025) · better on 4 of 7 shared results
- VideoComposer — VideoComposer: Compositional Video Synthesis with Motion Controllability (NeurIPS 2023) · better on 5 of 5 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
- FlowStyler: Artistic Video Stylization via Transformation Fields Transports (ICCV 2025 · similarity 0.90)
- StyleMaster: Stylize Your Video with Artistic Generation and Translation (CVPR 2025 · similarity 0.89)
- Generative Human Motion Stylization in Latent Space (ICLR 2024 · similarity 0.88)
- Encapsulated Composition of Text-to-Image and Text-to-Video Models for High-Quality Video Synthesis (CVPR 2025 · similarity 0.88)
- Conditional Balance: Improving Multi-Conditioning Trade-Offs in Image Generation (CVPR 2025 · similarity 0.87)
- Scalable Motion Style Transfer with Constrained Diffusion Generation (AAAI 2024 · similarity 0.87)
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.
