Comparison of Low Bitrate Quantizers for Encoding Swedish Sign Language

Anna Klezovich, Johanna Mesch, Gustav Eje Henter, Jonas Beskow


Abstract
This paper investigates the bitrate–distortion trade-off of different discrete representations for Swedish Sign Language (STS) using the STS Mocap v1 motion capture dataset. We compare the K-Means algorithm with the Residual Vector Quantized Variational Autoencoder (RQ-VAE) to determine how efficiently each method preserves salient motion information at low bitrates. The results show that RQ-VAE consistently achieves lower reconstruction error than K-Means at matching bitrates, particularly for body motion, and better preserves the signing space volume. We further demonstrate that quantized representations can serve as conditioning for a flow-matching generative model, producing plausible but still imperfect sign sequences at low bitrates. These findings highlight the advantages of vector quantized models for efficient sign language motion encoding.
Anthology ID:
2026.signlang-1.27
Volume:
Proceedings of the LREC 2026 12th Workshop on the Representation and Processing of Sign Languages: Language in Motion
Month:
May
Year:
2026
Address:
Palma, Mallorca (Spain)
Editors:
Eleni Efthimiou, Stavroula-Evita Fotinea, Thomas Hanke, Julie A. Hochgesang, Johanna Mesch, Marc Schulder
Venues:
SignLang | WS
SIG:
Publisher:
ELRA Language Resources Association (ELRA)
Note:
Pages:
256–261
Language:
External URL:
https://lrec.elra.info/lrec2026-ws-signlang-27
DOI:
10.63317/54ffsuydifyk
Bibkey:
Cite (ACL):
Anna Klezovich, Johanna Mesch, Gustav Eje Henter, and Jonas Beskow. 2026. Comparison of Low Bitrate Quantizers for Encoding Swedish Sign Language. In Proceedings of the LREC 2026 12th Workshop on the Representation and Processing of Sign Languages: Language in Motion, pages 256–261, Palma, Mallorca (Spain). ELRA Language Resources Association (ELRA).
Cite (Informal):
Comparison of Low Bitrate Quantizers for Encoding Swedish Sign Language (Klezovich et al., SignLang 2026)
Copy Citation: