Dance lesson generation with 3D motion capture data

Abstract

This project presents dance lesson generation method with 3D Motion Capture data. It targets on beginners and enable them to learn dancing through well organized dance lessons by themselves. Since dance learning is not just practice from the beginning to end, dance teachers have to segment and analyze dance manually. However, this task is time-consuming and students who self-learning dance may not know how to segment dance. This project aims to provide automatically dance lesson generation with imported digital audio file and motion capture data in order to provide a better dance self-learning method to stimulate the interests of students in dance learning. To achieve this, the technologies of beat detection, dance analysis and dance segmentation will be applied for the dance lesson generation. And, the structure of dance lessons is determined based on the complexity of the segmented motions.