Pose-Based Sign Language Appearance Transfer

Amit Moryossef, Gerard Sant, Zifan Jiang


Abstract
We introduce a method for transferring the signer’s appearance in sign language skeletal poses while preserving the sign content. Using estimated poses, we transfer the appearance of one signer to another, maintaining natural movements and transitions. This approach improves pose-based rendering and sign stitching while obfuscating identity. Our experiments show that while the method reduces signer identification accuracy, it slightly harms sign recognition performance, highlighting a tradeoff between privacy and utility.
Anthology ID:
2025.at4ssl-1.1
Volume:
Proceedings of the Third International Workshop on Automatic Translation for Signed and Spoken Languages (AT4SSL)
Month:
June
Year:
2025
Address:
Geneva, Switzerland
Editors:
Dimitar Shterionov, Mirella De Sisto, Bram Vanroy, Vincent Vandeghinste, Victoria Nyst, Myriam Vermeerbergen, Floris Roelofsen, Lisa Lepp, Irene Strasly
Venue:
AT4SSL
SIG:
Publisher:
European Association for Machine Translation
Note:
Pages:
1–6
Language:
URL:
https://preview.aclanthology.org/mtsummit-25-ingestion/2025.at4ssl-1.1/
DOI:
Bibkey:
Cite (ACL):
Amit Moryossef, Gerard Sant, and Zifan Jiang. 2025. Pose-Based Sign Language Appearance Transfer. In Proceedings of the Third International Workshop on Automatic Translation for Signed and Spoken Languages (AT4SSL), pages 1–6, Geneva, Switzerland. European Association for Machine Translation.
Cite (Informal):
Pose-Based Sign Language Appearance Transfer (Moryossef et al., AT4SSL 2025)
Copy Citation:
PDF:
https://preview.aclanthology.org/mtsummit-25-ingestion/2025.at4ssl-1.1.pdf