Application for the Recognition of Peruvian Sign Language using a Recurrent Neural Network

Research into effective communication between individuals with hearing disabilities and listeners remains a significant challenge that requires innovative technological solutions. This study focuses on the development of a mobile application leveraging a Recurrent Neural Network (RNN), specifically a Long Short-Term Memory (LSTM) model, to facilitate smooth communication of the Peruvian sign language. Initial experiments yielded promising results, with the system achieving 98% accuracy and a 3-second response time. Tests were conducted with 40 users, divided between those proficient in sign language and those unfamiliar with it. However, due to the lack of an existing Peruvian Sign Language dataset, a new dataset was created from scratch, consisting of 15 distinct gestures. Future iterations will need to address this limitation to optimize performance with a larger dataset.