Russian Fingerspelling Recognition Using Leap Motion Controller

Russian Sign Language is a national linguistic system with its vocabulary and grammar used by deaf people. In this paper, we focus on fingerspelling - a separate component of sign language. The developed system is capable of recognizing Russian fingerspelling static gestures captured with a Leap Motion Controller. In addition, we collected a new dataset of 400 static images of signs to evaluate the accuracy of the deep learning sign recognition system. The system achieved a success rate of 97.5%.