Real-Time Sign Language Translator: Bridging Communication Barriers with Mobile Machine Learning

Sign language is the primary form of communication in the deaf and hard-of-hearing community; however, this is not often understood by the general population. This paper details a real-time Sign Language Translator developed using Flutter and TensorFlow Lite. The system captures hand gestures with a smartphone camera and then processes them with a custom-trained machine learning model to translate real time gesture input to textual output. Our approach emphasizes accessibility, taking benefit of mobile technology to create a portable solution. The capability of the system was tested on different devices with an accuracy of 85% in controlled environments. Future work will be focusing on increasing the complexity of gestures, multilingual support & wearable technology integration.