Research map / 3D vision / 3D generation
Bailando: 3D Dance Generation by Actor-Critic GPT With Choreographic Memory
Li Siyao, Weijiang Yu, Tianpei Gu, Chunze Lin, Quan Wang, Chen Qian, Chen Change Loy, Ziwei Liu
Check 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 upper/lower half-body separation (ablation) | Pose VQ-VAE · AIST++ test set reconstruction | FID_k: 41.21 -> 28.23 | Improved |
| add a global velocity prediction branch (ablation) | Pose VQ-VAE · AIST++ test set reconstruction | FID_k: 70.95 -> 28.23 | Improved |
| add velocity-and-acceleration loss terms (ablation) | Pose VQ-VAE · AIST++ test set reconstruction | FID_k: 30.91 -> 28.23 | Improved |
| add velocity-and-acceleration loss terms (ablation) | Pose VQ-VAE · AIST++ test set reconstruction | FID_g: 11.87 -> 12.63 | Mixed |
| quantize dance poses into codebook tokens (ablation) | Motion GPT · AIST++ test set | FID_g: 147.28 -> 9.62 | Improved |
| use cross-conditional attention instead of separate causal attention (ablation) | Motion GPT · AIST++ test set | FID_k: 37.41 -> 28.16 | Improved |
| apply actor-critic finetuning with beat-align reward (ablation) | Motion GPT · AIST++ test set | BAS: 0.2245 -> 0.2332 | Improved |
| apply actor-critic finetuning with beat-align reward (ablation) | Motion GPT · AIST++ test set | FID_g: 11.82 -> 9.62 | Improved |
Later papers that compared with this method
5 later accepted papers report Bailando as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- OpenDance: Multimodal Controllable 3D Dance Generation with Large-scale Internet Data (CVPR 2026 · better on 1 of 2 shared results)
- InfiniteDance: Scalable 3D Dance Generation Towards in-the-wild Generalization (ECCV 2026 · better on 2 of 2 shared results)
- MEGADance: Mixture-of-Experts Architecture for Genre-Aware 3D Dance Generation (NeurIPS 2025 · better on 4 of 4 shared results)
- EgoMusic-driven Human Dance Motion Estimation with Skeleton Mamba (ICCV 2025 · better on 3 of 3 shared results)
- POPDG: Popular 3D Dance Generation with PopDanceSet (CVPR 2024 · better on 3 of 3 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
- Bailando ++: 3D Dance GPT With Choreographic Memory (TPAMI 2023 · similarity 0.98)
- DanceAnyWay: Synthesizing Beat-Guided 3D Dances with Randomized Temporal Contrastive Learning (AAAI 2024 · similarity 0.90)
- MEGADance: Mixture-of-Experts Architecture for Genre-Aware 3D Dance Generation (NeurIPS 2025 · similarity 0.89)
- Bidirectional Autoregessive Diffusion Model for Dance Generation (CVPR 2024 · similarity 0.88)
- Duolando: Follower GPT with Off-Policy Reinforcement Learning for Dance Accompaniment (ICLR 2024 · similarity 0.88)
- Explore 3D Dance Generation via Reward Model from Automatically-Ranked Demonstrations (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.
