Integrasi Model Pembelajaran Mesin dalam Game Menggunakan Gerakan Tangan

Abstract

This study develops a Tetris game controlled through hand gestures using a machine learning model. The primary objective of this research is to create an interactive and responsive gaming experience by utilizing hand gesture detection as the main control mechanism. A hand gesture dataset was collected from videos segmented into individual frames, which were then analyzed using MediaPipe to detect and label gestures. The machine learning model employs a Convolutional Neural Network (CNN) trained to recognize hand gesture patterns and translate them into commands within the game. After implementation, an evaluation was conducted by distributing questionnaires to 18 Informatics students at Adzkia University to assess the system's comfort and responsiveness. The questionnaire results showed a high satisfaction level, with an average score of 84.56, covering evaluations of control ease, gesture detection accuracy, and system responsiveness. The average score for ease of use reached 85, indicating that the majority of users found the gesture-based controls comfortable. This study demonstrates that applying machine learning models in gesture-based control games can provide a more interactive and responsive experience, with potential applications in other interactive technologies.