Sign language detection

Abstract

Communication is defined as the exchange of information, ideas and feelings.In order for two people to communicate with each other they must understand the same language.When a person is deaf they cannot hear and when they are mute they cannot speak.So in order to communicate with each other these groups of people use sign language.Being the community of the population this makes it difficult for them to communicate with the rest of the world. To overcome this barrier with the technologies that exist today we can create machine learning based models so that a deaf and dumb person with the help of this model can communicate with a normal person without knowing sign language.In this project we propose to create a dataset for Greek sign language phrase and word recognition so that real-time recognition can be performed.Real-time sign language recognition could solve many of the communication problems between deaf and non-deaf people.To achieve the desired result we train a model to create a real-time recognition system.The model reaches a high level of accuracy even though the data is not very accurate.