Research map / Video / Video generation
Beat-It: Beat-Synchronized Multi-Condition 3D Dance Generation
Zikai Huang, Xuemiao Xu, Cheng Xu, Huaidong Zhang, Chenxi Zheng, Jing Qin, Shengfeng He
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 hierarchical multi-condition fusion instead of one-stage concatenation (ablation) | Beat-It · AIST++ | PFC: 25.626 -> 0.966 | Improved |
| add hierarchical multi-condition fusion (ablation) | Beat-It · AIST++ | BAP: 0.323 -> 0.793 | Improved |
| add beat-aware dilation instead of vanilla masked attention (ablation) | Beat-It · AIST++ | BAP: 0.371 -> 0.793 | Improved |
| add beat-aware dilation instead of vanilla masked attention (ablation) | Beat-It · AIST++ | KPD: 0.389 -> 0.306 | Improved |
| add beat alignment loss (ablation) | Beat-It · AIST++ | BAS: 0.397 -> 0.661 | Improved |
| add beat alignment loss (ablation) | Beat-It · AIST++ | BAP: 0.411 -> 0.793 | Improved |
| use Beat-It with beat condition and keyframes instead of EDGE keyframe blending | dance generation · AIST++ | BAS: 0.235 -> 0.661 | Improved |
| use Beat-It instead of EDGE | dance generation · AIST++ | KPD: 0.859 -> 0.306 | Improved |
| use Beat-It instead of EDGE | dance generation · AIST++ | Div_k: 9.743 -> 9.660 | No clear effect |
Closest papers by meaning
- DanceAnyWay: Synthesizing Beat-Guided 3D Dances with Randomized Temporal Contrastive Learning (AAAI 2024 · similarity 0.91)
- Align Your Rhythm: Generating Highly Aligned Dance Poses with Gating-Enhanced Rhythm-Aware Feature Representation (ICCV 2025 · similarity 0.90)
- Music-Aligned Holistic 3D Dance Generation via Hierarchical Motion Modeling (ICCV 2025 · similarity 0.88)
- BRACE: The Breakdancing Competition Dataset for Dance Motion Synthesis (ECCV 2022 · similarity 0.88)
- OpenDance: Multimodal Controllable 3D Dance Generation with Large-scale Internet Data (CVPR 2026 · similarity 0.88)
- MEGADance: Mixture-of-Experts Architecture for Genre-Aware 3D Dance Generation (NeurIPS 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.
