Research map / Image generation / Autoregressive and latent generation
Dancing to Music
Hsin-Ying Lee, Xiaodong Yang, Ming-Yu Liu, Ting-Chun Wang, Yu-Ding Lu, Ming-Hsuan Yang, Jan Kautz
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 the composition phase to the decomposition-only model (ablation) | Music-to-dance generation · Testing music clips | FID: 14.8 -> 12.8 | Improved |
| Add the composition phase to the decomposition-only model (ablation) | Music-to-dance generation · Testing music clips | Beat Hit Rate: 72.4 % -> 65.1 % | Hurt |
| Add the composition phase to the decomposition-only model (ablation) | Music-to-dance generation · Testing music clips | Diversity: 49.7 ± 2.0 -> 53.2 ± 2.5 | Improved |
| Add the composition phase to the decomposition-only model (ablation) | Music-to-dance generation · Testing music clips | Multimodality: 51.4 ± 0.8 -> 47.8 ± 0.9 | Hurt |
| Add the composition phase to the decomposition-only model (ablation) | Music-to-dance generation · Testing music clips | Beat Coverage: 37.8 % -> 39.4 % | Improved |
| Use stochastic generation rather than deterministic LSTM mapping | Music-to-dance generation · Testing music clips | Multimodality | Improved |
| Decompose and temporally normalize dances into basic units, then compose them conditioned on music | Music-to-dance generation · Testing music clips | FID: 12.8 ± 0.8 vs 21.7 ± 0.8 | Improved |
| Use coherent composition rather than independent dance-unit concatenation (ablation) | Music-to-dance generation · Testing music clips | Motion realism preference: 61.2% | Improved |
| Use coherent composition rather than independent dance-unit concatenation (ablation) | Music-to-dance generation · Testing music clips | Style consistency preference: 68.3% | Improved |
Closest papers by meaning
- Dance Revolution: Long-Term Dance Generation with Music via Curriculum Learning (ICLR 2021 · similarity 0.92)
- Mining and Applying Composition Knowledge of Dance Moves for Style-Concentrated Dance Generation (AAAI 2023 · similarity 0.91)
- MNET++: Music-Driven Pluralistic Dancing Toward Multiple Dance Genre Synthesis (TPAMI 2023 · similarity 0.91)
- DanceAnyWay: Synthesizing Beat-Guided 3D Dances with Randomized Temporal Contrastive Learning (AAAI 2024 · similarity 0.90)
- Music-to-Dance Generation via Atomic Movements (ECCV 2026 · similarity 0.90)
- Bidirectional Autoregessive Diffusion Model for Dance Generation (CVPR 2024 · similarity 0.89)
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.
