A real time framework for AI basedHand gesture recognition

Abstract

The normal people on the earth used to talk to each other using different languages but in case of dumb and hearing impaired, they used to communicate with each other through a standard sign language represented using their signs and hand-movement. But the communication between a Dumb& Deafpersonand a normal person who do not know that sign language, it becomes inexplicably difficult. To avoid that scenario an intermediate device called smart translator can serve the purpose which is the motivation. The entire work aims at constructing a standard algorithm that detects the hand movements and treats them as digits which are mapped with English characters represented using Standard sign language.Suppose if it needs to represent a word “Covid” using sign language, a set of 5 different hand movements are to be calibrated individually to complete the sequence of characters to form the word  and likewise, the sentences and as well as paragraphs can be framed.Then the translation of speech from that obtained string can be done to complete the communication. The primary objective of this work is to construct a medium between normal and speech impaired person who uses sign language by using hand gestures. Efforts were made to set the model cost efficient to be incorporated into a final real time frame work.