The Development and Testing of a Model for Converting American Sign Language (ASL) Sentences Using Natural Language Processing Techniques

This research addresses the global prevalence of people with hearing disabilities, recognizing the increasing numbers within this demographic. The communication challenges encountered by these individuals drive the development and evaluation of a language translation model. The primary goal is to design a model specifically tailored for translating sign language sentences into English, considering the unique word order in sign language. The study involves comparing eight models that utilize various techniques. Evaluation will be conducted through the bleu score, with 100 randomly selected sentences in each of the 10 rounds. Consistent inputs will be used across rounds to calculate the bleu score, offering a comprehensive assessment of model performance.