Saved in:
Bibliographic Details
Main Authors: Yao, Huijie, Zhou, Wengang, Zhou, Hao, Li, Houqiang
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.08173
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916283687108608
author Yao, Huijie
Zhou, Wengang
Zhou, Hao
Li, Houqiang
author_facet Yao, Huijie
Zhou, Wengang
Zhou, Hao
Li, Houqiang
contents Spoken language glossification (SLG) aims to translate the spoken language text into the sign language gloss, i.e., a written record of sign language. In this work, we present a framework named $S$emi-$S$upervised $S$poken $L$anguage $G$lossification ($S^3$LG) for SLG. To tackle the bottleneck of limited parallel data in SLG, our $S^3$LG incorporates large-scale monolingual spoken language text into SLG training. The proposed framework follows the self-training structure that iteratively annotates and learns from pseudo labels. Considering the lexical similarity and syntactic difference between sign language and spoken language, our $S^3$LG adopts both the rule-based heuristic and model-based approach for auto-annotation. During training, we randomly mix these complementary synthetic datasets and mark their differences with a special token. As the synthetic data may be less quality, the $S^3$LG further leverages consistency regularization to reduce the negative impact of noise in the synthetic data. Extensive experiments are conducted on public benchmarks to demonstrate the effectiveness of the $S^3$LG. Our code is available at \url{https://github.com/yaohj11/S3LG}.
format Preprint
id arxiv_https___arxiv_org_abs_2406_08173
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Semi-Supervised Spoken Language Glossification
Yao, Huijie
Zhou, Wengang
Zhou, Hao
Li, Houqiang
Computation and Language
Spoken language glossification (SLG) aims to translate the spoken language text into the sign language gloss, i.e., a written record of sign language. In this work, we present a framework named $S$emi-$S$upervised $S$poken $L$anguage $G$lossification ($S^3$LG) for SLG. To tackle the bottleneck of limited parallel data in SLG, our $S^3$LG incorporates large-scale monolingual spoken language text into SLG training. The proposed framework follows the self-training structure that iteratively annotates and learns from pseudo labels. Considering the lexical similarity and syntactic difference between sign language and spoken language, our $S^3$LG adopts both the rule-based heuristic and model-based approach for auto-annotation. During training, we randomly mix these complementary synthetic datasets and mark their differences with a special token. As the synthetic data may be less quality, the $S^3$LG further leverages consistency regularization to reduce the negative impact of noise in the synthetic data. Extensive experiments are conducted on public benchmarks to demonstrate the effectiveness of the $S^3$LG. Our code is available at \url{https://github.com/yaohj11/S3LG}.
title Semi-Supervised Spoken Language Glossification
topic Computation and Language
url https://arxiv.org/abs/2406.08173