A Novel FACS-Aligned Anatomical Text Description Paradigm for Fine-Grained Facial Behavior Synthesis
Jiahe Wang, Cong Liang, Xuandong Huang, Yuxin Wang, Xin Yun, Yi Wu, Yanan Chang, Shangfei Wang · 2026
Read A Novel FACS-Aligned Anatomical Text Description Paradigm for Fine-Grained Facial Behavior Synthesis by Jiahe Wang, Cong Liang, Xuandong Huang, Yuxin Wang, Xin Yun, Yi Wu, Yanan Chang, Shangfei Wang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-18588.txt.
Description
Wang and co-authors turn facial action coding into rich text prompts for face synthesis. Coarse emotion labels and one-hot action unit vectors often miss fine muscle moves. Their method writes FACS muscle rules and conflict fixes into plain text signals. A dynamic processor converts raw action unit notes into sound descriptions. The result is BP4D-AUText, over 302,000 paired text and image samples. Their VQ-AUFace model beats prior methods most clearly when action units conflict.
Catalog record
| CREATOR | Jiahe Wang, Cong Liang, Xuandong Huang, Yuxin Wang, Xin Yun, Yi Wu, Yanan Chang, Shangfei Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 168325 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.18588