Multilingual Gloss-free Sign Language Translation: Towards Building a Sign Language Foundation Model

TL;DR

This paper introduces a multilingual, gloss-free sign language translation model that supports multiple SLs and various translation tasks, improving accessibility and performance across diverse benchmarks.

Contribution

It presents a novel dual CTC-based multilingual model for sign language translation without gloss annotations, handling multiple SLs and translation scenarios.

Findings

Supports 10 sign languages and multiple translation tasks.

Achieves competitive results on three benchmark datasets.

Addresses low-resource and language conflict challenges in multilingual SLT.

Abstract

Sign Language Translation (SLT) aims to convert sign language (SL) videos into spoken language text, thereby bridging the communication gap between the sign and the spoken community. While most existing works focus on translating a single sign language into a single spoken language (one-to-one SLT), leveraging multilingual resources could mitigate low-resource issues and enhance accessibility. However, multilingual SLT (MLSLT) remains unexplored due to language conflicts and alignment difficulties across SLs and spoken languages. To address these challenges, we propose a multilingual gloss-free model with dual CTC objectives for token-level SL identification and spoken text generation. Our model supports 10 SLs and handles one-to-one, many-to-one, and many-to-many SLT tasks, achieving competitive performance compared to state-of-the-art methods on three widely adopted benchmarks: multilingual SP-10, PHOENIX14T, and CSL-Daily.