Real-Time Sign Language Translator for Children with Hearing Impairments Using Gaussian Naive Bayes Algorithm and Mediapipe Fadly Shabir, Arysespajayadi, Ahmad Irfan Abdullah and Muh. Aidil
Politeknik Politeknik Negeri Media Kreatif, Jl. Perintis Kemerdekaan VI No.50, Tamalanrea Jaya, Kec. Tamalanrea, Kota Makassar, Sulawesi Selatan 90245
Abstract
This research aims to develop a sign language translator application for deaf children. The system utilizes MediaPipe and applies the Gaussian Naive Bayes Algorithm. The primary focus of the study is to enhance communication skills among hearing-impaired children through the interpretation of sign language gestures. The research is structured into several stages. Initially, sign language gesture data is gathered from various sources. Subsequently, the data is analysed to identify relevant gestures. The dataset is then split into two parts: training data for model development and experimental data for evaluation. Next, features are extracted from the gestures, and a Gaussian Naive Bayes model is constructed using the training data. Following the training process, the model is tested using the experimental data to assess system performance. The aim is for this research to make a significant contribution by improving communication abilities for children with hearing impairments through the development of an efficient and accurate system or application for translating hand gestures.
