Neural Network Based Method for Sign Language Recognition

Abstract

Sign language is a language that especially abled people use to communicate with normal people. Speech impairment is a disability which affects one’s ability to speak and hear. Hand signs is one of the methods used for non-verbal communications which we call Sign Language. The goal of this project is to provide a simple way to very trivial communication problems the world faces in this present scenario. We aim to build a minimalistic translation engine that recognizes and converts sign language to text (English) and could be implemented on any platform. This engine could be made with the help of Image Processing and Deep Learning.

 Keywords: OpenCV, Darknet YOLOv3, Image Recognition, Convolutional Neural Network, Sign Language, Machine Learning, Supervised Learning.