Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.03799 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911722399334400 |
|---|---|
| author | Zheng, Zhiwei Jin, Shibo Liu, Lingjie Zhao, Mingmin |
| author_facet | Zheng, Zhiwei Jin, Shibo Liu, Lingjie Zhao, Mingmin |
| contents | Autoregressive (AR) models offer stable and efficient training, but standard next-token prediction is not well aligned with the temporal structure required for text-conditioned motion generation. We introduce MoScale, a next-scale AR framework that generates motion hierarchically from coarse to fine temporal resolutions. By providing global semantics at the coarsest scale and refining them progressively, MoScale establishes a causal hierarchy better suited for long-range motion structure. To improve robustness under limited text-motion data, we further incorporate cross-scale hierarchical refinement for improving per-scale initial predictions and in-scale temporal refinement for selective bidirectional re-prediction. MoScale achieves SOTA text-to-motion performance with high training efficiency, scales effectively with model size, and generalizes zero-shot to diverse motion generation and editing tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_03799 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Next-Scale Autoregressive Models for Text-to-Motion Generation Zheng, Zhiwei Jin, Shibo Liu, Lingjie Zhao, Mingmin Computer Vision and Pattern Recognition Autoregressive (AR) models offer stable and efficient training, but standard next-token prediction is not well aligned with the temporal structure required for text-conditioned motion generation. We introduce MoScale, a next-scale AR framework that generates motion hierarchically from coarse to fine temporal resolutions. By providing global semantics at the coarsest scale and refining them progressively, MoScale establishes a causal hierarchy better suited for long-range motion structure. To improve robustness under limited text-motion data, we further incorporate cross-scale hierarchical refinement for improving per-scale initial predictions and in-scale temporal refinement for selective bidirectional re-prediction. MoScale achieves SOTA text-to-motion performance with high training efficiency, scales effectively with model size, and generalizes zero-shot to diverse motion generation and editing tasks. |
| title | Next-Scale Autoregressive Models for Text-to-Motion Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.03799 |