Unraveling a Decade: A Comprehensive Survey on Isolated Sign Language Recognition

Sign language plays a crucial role as a distinct and vital mode of communication for diverse groups of people in society. Each sign language encompasses a wide array of signs, each characterized by unique local and global articulations, e.g. hand shape, motion profile, and the arrangement of the hands, face, and body. Consequently, the domain of visual Sign Language Recognition (SLR) presents a complex and challenging research area within the field of computer vision, even with state-of-the-art models. This survey paper provides a comprehensive overview of Isolated Sign Language Recognition (ISLR), covering various aspects including input modality, modelled sign language parameters, fusion methods, and transfer learning, all of which have an impact on the performance of SLR methods. In addition, we present an overview of publicly available benchmark datasets for ISLR as well as analyze the state-of-the-art results achieved on these datasets. By examining these different aspects along with benchmarking strategies, we provide insights into the advancements, challenges, and potential directions in ISLR research.