Applying Machine Learning for American Sign Language Recognition: A Brief Survey

Abstract

Sign languages play a crucial role in enabling differently-abled people to communicate and express their feelings. The advent of newer technologies in machine learning and sensors has led us to build more sophisticated and complex human-computer interfaces. However, a cost-effective commercial sign language recognition system is still not available. In this paper, we have highlighted recent advancements and limitations related to sign language recognition systems. We have focused our study on American sign language (ASL), which is one of the most widely used sign languages. We have tried to classify most of the surveyed work into two broad areas based on data acquisition techniques: visual and sensor-based approaches. We have also listed publicly available datasets that could be used for further research. This paper aims to provide a clear insight into advancement in the field of sign language recognition systems using American sign language. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.