A Study on Techniques and Challenges in Sign Language Translation

Abstract

Sign Language Translation (SLT) plays a pivotal role in enabling effective communication for the Deaf and Hard of Hearing (DHH) community. This review delves into the state-of-the-art techniques and methodologies in SLT, focusing on its significance, challenges, and recent advancements. The review provides a comprehensive analysis of various SLT approaches, ranging from rule-based systems to deep learning models, highlighting their strengths and limitations. Datasets specifically tailored for SLT research are explored, shedding light on the diversity and complexity of Sign Languages across the globe. The review also addresses critical issues in SLT, such as the expressiveness of generated signs, facial expressions, and non-manual signals. Furthermore, it discusses the integration of SLT into assistive technologies and educational tools, emphasizing the transformative potential in enhancing accessibility and inclusivity. Finally, the review outlines future directions, including the incorporation of multimodal inputs and the imperative need for co-creation with the Deaf community, paving the way for more accurate, expressive, and culturally sensitive Sign Language Generation systems.