SignBank+: Preparing a Multilingual Sign Language Dataset for Machine Translation Using Large Language Models

Amit Moryossef, Zifan Jiang
9/20/2023
cs.CL

No official code found. Best implementation: sign-language-processing/signbank-plus (22 stars, last commit 2024-02-22).

Abstract

We introduce SignBank+, a clean version of the SignBank dataset, optimized for machine translation between spoken language text and SignWriting, a phonetic sign language writing system. In addition to previous work that employs complex factorization techniques to enable translation between text and SignWriting, we show that a traditional text-to-text translation approach performs equally effectively on the cleaned SignBank+ dataset. Our evaluation results indicate that models trained on SignBank+ surpass those on the original dataset, establishing a new benchmark for SignWriting-based sign language translation and providing an open resource for future research.

View on arXivView PDF

Code Implementations(1)

sign-language-processing/signbank-plusCommunity100%

Sign language translation dataset using SignWriting

226Shell, PythonSep 20, 20232 years ago

Similar Papers

Saudi Sign Language Translation Using T5

92% match2025

DualAnchor: Preserving Language Priors and Improving Lexical Fidelity in Gloss-Free Sign Language Translation

89% match2026

Sign Language Translation from Instructional Videos

89% match1 repo512023

Sign2GPT: Leveraging Large Language Models for Gloss-Free Sign Language Translation

88% match1 repo462026

Direct Translation between Sign Languages

88% match2026

RVLF: A Reinforcing Vision-Language Framework for Gloss-Free Sign Language Translation

88% match2025

Cite this paper

@article{moryossef2023signbank,
  title  = {SignBank+: Preparing a Multilingual Sign Language Dataset for Machine Translation Using Large Language Models},
  author = {Amit Moryossef and Zifan Jiang},
  year   = {2023},
  eprint = {2309.11566},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2309.11566v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2309.11566.svg)](https://codeofpaper.com/paper/2309.11566)

Discussion