Research map / Video / Video generation
Hierarchical Style-based Networks for Motion Synthesis
Jingwei Xu, Huazhe Xu, Bingbing Ni, Xiaokang Yang, Xiaolong Wang, Trevor Darrell
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 long-range generation model on top of short-range sequences (ablation) | Hierarchical style-based networks · CMU Mocap motion diversity | motion deviation: 0.226 -> 0.347 | Improved |
| add the feature-level style loss L_trn during training (ablation) | Short-range style transfer model · Style feature consistency (one sequence) | style feature drift from reference | Improved |
| use our model instead of DeepSyn for generation | Reality classifier trained on QuaterNet negatives · Reality evaluation | proportion classified as positive: 41.9% -> 75.3% | Improved |
| use our model instead of QuaterNet | Reality classifier trained on DeepSyn negatives · Reality evaluation | proportion classified as positive: 37.6% -> 66.3% | Improved |
| use our model instead of DeepSyn | Distribution similarity · CMU Mocap | FVD-style distance: 281.5 -> 179.8 | Improved |
| use our model instead of QuaterNet | Distribution similarity · CMU Mocap | FVD-style distance: 341.1 -> 179.8 | Improved |
| increase training data from 10% to 100% | Our model diversity versus DeepSyn · CMU Mocap diversity | joint deviation | Improved |
Closest papers by meaning
- How to Move Your Dragon: Text-to-Motion Synthesis for Large-Vocabulary Objects (ICML 2025 · similarity 0.88)
- Generative Human Motion Stylization in Latent Space (ICLR 2024 · similarity 0.87)
- Scaling Large Motion Models with Million-Level Human Motions (ICML 2025 · similarity 0.87)
- OmniMotionGPT: Animal Motion Generation with Limited Data (CVPR 2024 · similarity 0.87)
- MotionDreamer: One-to-Many Motion Synthesis with Localized Generative Masked Transformer (ICLR 2025 · similarity 0.87)
- Single Motion Diffusion (ICLR 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.
