Towards AI-driven Sign Language Generation with Non-manual Markers

TL;DR

This paper presents an AI system that translates English into American Sign Language videos, incorporating facial cues and body language for more natural communication, based on recent advances in language and video generation.

Contribution

It introduces a novel AI-driven sign language generation system that effectively captures non-manual markers and improves translation quality over existing methods.

Findings

User study with 30 DHH participants shows high satisfaction.

System achieves more natural and accurate sign language videos.

Identifies key areas for further improvement in sign language synthesis.

Abstract

Sign languages are essential for the Deaf and Hard-of-Hearing (DHH) community. Sign language generation systems have the potential to support communication by translating from written languages, such as English, into signed videos. However, current systems often fail to meet user needs due to poor translation of grammatical structures, the absence of facial cues and body language, and insufficient visual and motion fidelity. We address these challenges by building on recent advances in LLMs and video generation models to translate English sentences into natural-looking AI ASL signers. The text component of our model extracts information for manual and non-manual components of ASL, which are used to synthesize skeletal pose sequences and corresponding video frames. Our findings from a user study with 30 DHH participants and thorough technical evaluations demonstrate significant progress and identify critical areas necessary to meet user needs.