K-NN for Hand Keypoints Detection Based on SIBI Standard

Indonesian Sign Language System (SIBI) is one of the sign languages used by deaf-impaired people to communicate in Indonesia. However, people generally still do not learn the SIBI language, which becomes an obstacle for deaf-impaired people to communicate with the community. Media is needed to translate the SIBI language to solve this problem. In this study, we design a translator system using the K-Nearest Neighbor (K-NN) classification method to translate the SIBI language. We will use the hand keypoints detection feature to simplify the hand detection process. Before building the system, the training and testing process is first carried out. The training process uses the K-Folds method with ten folds. From the training and testing process, the accuracy results of 93.70% and 93.75% with the K variable are 1. At the system testing stage, the accuracy results are 89.58%. This system aims to bridge the communication gap between deaf-impaired individuals and the broader community, facilitating better integration and understanding.