Indian Sign Language Recognition Using Random Forest Classifier
S Ajay, Ajith Potluri, Sara Mohan George, Ramashish Gaurav, S Anusri
Abstract
Communication is the foundation of all human relationships, both personal and professional. It is one of the basic requirements for survival in a society. Verbal communication is impossible without a well-defined language that is understood by both parties. Around 26% of the disabled population in India use sign language for communication. As a result, there is a pressing need to bridge the communication gap between the general public and the speech impaired. The proposed idea is to develop a pair of sensor gloves which detects Indian Sign Language (ISL) gestures and converts them into audible speech. The gloves are mounted with various sensors and modules such as the Arduino Nano microcontrollers, flex sensors, touch sensors, Inertial Measurement Units (IMU), RF and Bluetooth modules. With these sensors, the state of both hands can be quantified in a series of numerical data to capture their motion. The sensor values are also sent to a Machine Learning classification algorithm which improves the accuracy of gesture recognition. The audible speech output is obtained by creating an app on a smart phone that can convert text to speech.