Real-Time Sign Language Detection Using Deep Learning

Sign language is the primary mode of communication for deaf and hard of hearing community. This paper discusses the development of a real time sign language recognition system by deploying deep learning techniques. For instance, the aim is to establish a good model which could be used to identify different hand gestures as used in sign languages. Once trained on a data set of 10 different hand gestures, the system is able to recognize and predict gestures in other images. The use of TensorFlow and Keras in the implementation ensures scalability of the system for real-time application. This method aims at assisting the deaf and mute people by developing an application that will improve their social life. Furthermore, TensorFlow and Python facilitate the porting of the system to mobile and embedded systems, which enhance its reach among different gadgets. This system is intended to help avoid one of the main concerns of sign language users in which they have limited ability to interact with others and thus limited engagement with basic services. The objective of the researcher is to make the quality of life of sign language users better by translating their gestures into a format that is easier to understand.