Research map / Video / Video generation
PoseCrafter: One-Shot Personalized Video Synthesis Following Flexible Pose Control
Yong Zhong, Min Zhao, Zebin You, Xiaofeng Yu, Changwang Zhang, Chongxuan Li
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 reference-frame selection to ControlVideo (ablation) | ControlVideo · TED | FVD: 109.36 -> 74.51 | Improved |
| add reference-frame insertion into inference poses (ablation) | ControlVideo + selection · TED | FID: 22.03 -> 18.46 | Improved |
| add latent editing (affine transform of face and hand latents) (ablation) | PoseCrafter · TED | PSNR: 23.78 -> 23.92 | Improved |
| add latent editing (ablation) | PoseCrafter · TED | FVD: 66.91 -> 80.01 | Hurt |
| add latent editing (ablation) | PoseCrafter · TED | CLIP-T: 0.949 -> 0.954 | Improved |
| increase number of training frames from 8 to 32 (ablation) | PoseCrafter · TikTok | FVD: 440.49 -> 362.09 | Improved |
| increase training frames far beyond a point (ablation) | PoseCrafter · open-domain videos | PSNR | No clear effect |
| fine-tune Disco on a single video | Disco · TikTok | FVD: 689.23 -> 807.35 | Hurt |
| apply latent editing later in sampling (alpha from 1 to 50) (ablation) | PoseCrafter · open-domain videos | CLIP-T / SSIM | Hurt |
| use PoseCrafter (8 frames) instead of MagicAnimate on TikTok PSNR/FVD | MagicAnimate · TikTok | PSNR / FVD: PSNR 17.95 vs 17.36; FVD 374.40 vs 440.49 | 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.
- ControlVideo — ControlVideo: Training-free Controllable Text-to-video Generation (ICLR 2024) · better on 8 of 8 shared results
- MagicAnimate — MagicAnimate: Temporally Consistent Human Image Animation using Diffusion Model (CVPR 2024) · better on 4 of 7 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
- HumanDreamer: Generating Controllable Human-Motion Videos via Decoupled Generation (CVPR 2025 · similarity 0.88)
- FaithfulFaces: Pose-Faithful Facial Identity Preservation for Text-to-Video Generation (NeurIPS 2026 · similarity 0.86)
- VLOGGER: Multimodal Diffusion for Embodied Avatar Synthesis (CVPR 2025 · similarity 0.86)
- ReMask-Animate: Refined Character Image Animation Using Mask-Guided Adapters (AAAI 2025 · similarity 0.86)
- Boximator: Generating Rich and Controllable Motions for Video Synthesis (ICML 2024 · similarity 0.86)
- Follow Your Pose: Pose-Guided Text-to-Video Generation Using Pose-Free Videos (AAAI 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.
