YouTube-SL-25: A Large-Scale, Open-Domain Multilingual Sign Language Parallel Corpus

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tanzer, Garrett, Zhang, Biao
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929421897695232
author Tanzer, Garrett
Zhang, Biao
author_facet Tanzer, Garrett
Zhang, Biao
contents Even for better-studied sign languages like American Sign Language (ASL), data is the bottleneck for machine learning research. The situation is worse yet for the many other sign languages used by Deaf/Hard of Hearing communities around the world. In this paper, we present YouTube-SL-25, a large-scale, open-domain multilingual corpus of sign language videos with seemingly well-aligned captions drawn from YouTube. With >3000 hours of videos across >25 sign languages, YouTube-SL-25 is a) >3x the size of YouTube-ASL, b) the largest parallel sign language dataset to date, and c) the first or largest parallel dataset for many of its component languages. We provide baselines for sign-to-text tasks using a unified multilingual multitask model based on T5 and report scores on benchmarks across 4 sign languages. The results demonstrate that multilingual transfer benefits both higher- and lower-resource sign languages within YouTube-SL-25.
format Preprint
id arxiv_https___arxiv_org_abs_2407_11144
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle YouTube-SL-25: A Large-Scale, Open-Domain Multilingual Sign Language Parallel Corpus
Tanzer, Garrett
Zhang, Biao
Computation and Language
Even for better-studied sign languages like American Sign Language (ASL), data is the bottleneck for machine learning research. The situation is worse yet for the many other sign languages used by Deaf/Hard of Hearing communities around the world. In this paper, we present YouTube-SL-25, a large-scale, open-domain multilingual corpus of sign language videos with seemingly well-aligned captions drawn from YouTube. With >3000 hours of videos across >25 sign languages, YouTube-SL-25 is a) >3x the size of YouTube-ASL, b) the largest parallel sign language dataset to date, and c) the first or largest parallel dataset for many of its component languages. We provide baselines for sign-to-text tasks using a unified multilingual multitask model based on T5 and report scores on benchmarks across 4 sign languages. The results demonstrate that multilingual transfer benefits both higher- and lower-resource sign languages within YouTube-SL-25.
title YouTube-SL-25: A Large-Scale, Open-Domain Multilingual Sign Language Parallel Corpus
topic Computation and Language
url https://arxiv.org/abs/2407.11144