Sign Language Translation Using Machine Learning And TensorFlow

Research Supervisor
Journal Title
Journal Title
Journal ISSN
Volume Title
Sign Language play a crucial role in nonverbal communication and essential to daily living. Sign Language Recognition, which has been researched for many years, is a breakthrough for assisting deaf-mute people. Unfortunately, each study has its own limitations and cannot be used commercially. Some studies have proven to be successful in recognizing sign language, but commercialization is prohibitively expensive.With the aid of a hand gesture detection system, we have access to a novel, comfortable, and user-friendly method of interacting with computers that is more suited to human needs. We intend to redefine the Sign Language Translator using deep learning. Our aim is to make an improved model that will come close to or even improve on current models. Our first task was to understand existing projects and models, which is why we were comparing our own findings with the findings of current research papers. This will inspire us to take the best parts of the existing models and put our own spin on them in the future. Even though we had a very rocky start, getting very poor results compared to our target, we managed to increase our accuracy by learning more and evolving our findings and understanding of the topic. We concluded with improved results and are currently at half way point in our project. In the next half, we intend to make the model using CNN and also make a recognizer to recognize the signs.
Department Name
Electrical and Computer Engineering
North South University
Printed Thesis