Research map / Video / Video generation
FreeDance: Towards Harmonic Free-Number Group Dance Generation via a Unified Framework
Yiwen Zhao, Yang Wang, Liting Wen, Hengyuan Zhang, Xingqun Qi
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 2D token map instead of 1D codebook (ablation) | FreeDance stage one · AIST++ & AIOZ-GDANCE (mixed) | FID: 6.62 -> 4.46 | Improved |
| use 2D token map instead of 1D codebook (ablation) | FreeDance stage one · AIST++ & AIOZ-GDANCE (mixed) | Diversity: 2.06 -> 2.70 | Improved |
| add person-number-dependent quantization (partition) (ablation) | FreeDance 2D codebook · AIST++ & AIOZ-GDANCE (mixed) | FID: 4.46 -> 1.15 | Improved |
| add person-number-dependent quantization (partition) (ablation) | FreeDance 2D codebook · AIST++ & AIOZ-GDANCE (mixed) | BAS: 0.2433 -> 0.2609 | Improved |
| add person-number-dependent quantization (partition) (ablation) | FreeDance 2D codebook · AIST++ & AIOZ-GDANCE (mixed) | Diversity: 2.70 -> 3.42 | Improved |
| add CRAM (ablation) | FreeDance masked token modeling · AIST++ & AIOZ-GDANCE (mixed) | BAS: 0.2411 -> 0.2569 | Improved |
| add CRAM (ablation) | FreeDance masked token modeling · AIST++ & AIOZ-GDANCE (mixed) | FID: 2.24 -> 2.08 | Improved |
| add TIM (ablation) | FreeDance masked token modeling · AIST++ & AIOZ-GDANCE (mixed) | Diversity: 2.70 -> 3.26 | Improved |
| add TIM (ablation) | FreeDance masked token modeling · AIST++ & AIOZ-GDANCE (mixed) | FID: 2.24 -> 2.06 | Improved |
| add both CRAM and TIM (ablation) | FreeDance masked token modeling · AIST++ & AIOZ-GDANCE (mixed) | FID: 2.24 -> 1.15 | Improved |
| add both CRAM and TIM (ablation) | FreeDance masked token modeling · AIST++ & AIOZ-GDANCE (mixed) | BAS: 0.2411 -> 0.2609 | Improved |
| add both CRAM and TIM (ablation) | FreeDance masked token modeling · AIST++ & AIOZ-GDANCE (mixed) | Diversity: 2.70 -> 3.42 | Improved |
Closest papers by meaning
- Breaking the Group Size Barrier: Parameter-Efficient Group Dance Generation with Chain-of-Dancers (NeurIPS 2026 · similarity 0.90)
- Harmonious Music-driven Group Choreography with Trajectory-Controllable Diffusion (AAAI 2025 · similarity 0.88)
- DanceAnyWay: Synthesizing Beat-Guided 3D Dances with Randomized Temporal Contrastive Learning (AAAI 2024 · similarity 0.87)
- Duolando: Follower GPT with Off-Policy Reinforcement Learning for Dance Accompaniment (ICLR 2024 · similarity 0.87)
- MEGADance: Mixture-of-Experts Architecture for Genre-Aware 3D Dance Generation (NeurIPS 2025 · similarity 0.86)
- DisCo: Disentangled Control for Realistic Human Dance Generation (CVPR 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.
