SIGN LANGUAGE RECOGNITION USING DEEP LEARNING









Abstract

This paper proposes an approach to solve the problem of decoding the sign language which goes by when a person is talking in sign language in coverage of the web camera used in our system, then a corresponding meaning associated to the gesture is displayed on the computer screen along the percentage efficiency of that gesture. We previously were focused on working sign language but all the available approaches were outdated and therefore by making use of new and recent technologies we developed gesture detection system which can also serve as sign language translator and also there is a speech conversion feature added in this project. This system makes use of the following dependencies which are OpenCV, TensorFlow API, and Keras. The main purpose of our system is to reduce the loss and increase the efficiency of hand gestures in day-to-day life. The existing system and the modified system are both present in this project.


Modules


Algorithms


Software And Hardware