Sign Language Interpretation Using Deep Learning

Sign language is used to communicate a particular message over some universally known and accepted gestures. The speech and hearing challenged people use a specific combination of hand gestures and movements to convey a message. Despite the extensive research progress in Sign Language Detection, cost effective and performance effective solutions are still need of the day. Deep learning, and computer vision can be used to provide an effective solution to the user. This can be very helpful for the hearing and speech impaired people in seamless communication with others around as knowing sign language is not something that is common to all. In this work, a sign detector is developed, which detects various signs of the Sign Language used by speech impaired people. Here, data taken as input in the form of images is extensively used for both training and testing using machine learning. A custom Convolutional Neural Network (CNN) model to identify the sign from an image frame using Open-CV is developed and sentence construction of the detected signs is accomplished. A lot of images have been used as input for the purposes of training and testing. Many of the symbols in sign language could be rightly identified. A series of gestures are translated as text to the recipient.