Real-Time Sign Language Identification using KNN: A Machine Learning Approach
Shrikant V. Sonekar, Harsh Dhoke, Vaibhav Mate, Snehal Dhewle, Marshneel Patil · 2023
A disability may actually constitute incapacity. Speech impairment is a handicap that affects a person’s capacity to communicate verbally and auditorily. People with this handicap use signing as their primary form of communication. A real-time sign detector could be a big improvement in how the hearing impaired and the general public communicate. We have created a solid model that reliably categorizes signing in the vast majority of situations involving several languages. Additionally, this approach is very helpful to those learning sign language in the interim. Throughout the study, various human-computer interaction approaches for gesture recognition were investigated and evaluated. By identifying a person’s hand signs and translating them into legible text, our initiative attempts to communicate with non-sign language users. We have accomplished this using the aid of OpenCV and machine learning, particularly KNN. By using an internet camera or phone camera to take a photo as input, we hope to create a model that can predict the sign and produce the output in form of text/ audio. The system is prepared to have the highest level of accuracy ever seen in a model of this kind.