Recent progress in sign language recognition: a review

Sign language is a predominant form of communication among a large group of society. The nature of sign languages is visual, making them distinct from spoken languages. Unfortunately, very few able people can understand sign language making communication with the hearing-impaired infeasible. Research in the field of sign language recognition (SLR) can help reduce the barrier between deaf and able people. Despite having tremendous advances in SLR, unfortunately, this form of recognition is still at least a decade behind speech recognition. There has been a gradual transition from static to isolated to continuous SLR, but still the research is scattered, limited to very small vocabularies, and only suitable for tailor-made conditions. This paper aims to compile recent progress in SLR and presents a comprehensive review of the emerging SLR frameworks and algorithms. We have categorized SLR based on the unit of written text, i.e., letters or alphabets, words and sentences. This review also includes a study-wise summary of the datasets used in different research conducted during the last few years. We identify state-of-the-art techniques for each category. We also suggest novel research directions for future work, and highlight several primary factors contributing to SLR’s inability to achieve improved practical outcomes.