<![CDATA[Currently, semaphore learning is still done manually with instructor guidance, which has limitations in instructor availability and time efficiency. This research aims to create a semaphore game that implements PoseNet to increase the efficiency and interactivity of learning semaphore gestures. Through the application of PoseNet technology, this game aims to recognize the user's body movements in real-time. By identifying 17 key points on the human body, PoseNet enables automatic, real-time detection of semaphore gestures. The system development method and model used is the prototype & MDLC model, and the implementation of the p5.js and ml5.js libraries provides the basis for the integration of PoseNet into the Semaphore game. The results of this research present a semaphore game that implements PoseNet as an innovative solution to improve semaphore gesture learning.]]>
Copyrights © 2024
