AI-Driven Motion Capture for Silat Movement Animation

Abstract

Artificial Intelligence (AI) motion capture is revolutionizing the method of capturing and animating virtual characters or objects, greatly improving its accessibility for users. It provides adaptability by accommodating various environments, facilitating real-time functionalities with immediate feedback for active applications, and ensuring user-friendly procedures. In this paper, AI-driven motion capture to record and translate the intricate movements of Silat into animations to bring this traditional martial art to life in the digital realm can enhance the game's authenticity and immersion. Silat is a martial art also known as pencak silat. However, the remaining problem is that some people may find it challenging to follow traditional methods of studying Silat since they can be expensive and demand a lot of time and dedication. Furthermore, a lot of the current Silat learning resources are static and lack interaction, which makes it difficult for learners to fully engage with the content. Therefore, this study aims to develop a Silat animation based on AI motion capture for virtual heritage. Three phases have been carried out. First is studying and analyzing the existing 3D exploration game for virtual heritage, and the basic movement of Silat animation. The second phase is reconstructing the motion-based Silat using AI. Finally, the third phase is to measure the user experience for learning Silat with the precise movement produced by AI motion capture.