Research and Implementation of Conductors' Beat Recognition System

• Main Authors: Tung-Ping Wang, 王東平
• Other Authors: Jyh-Shing Roger Jang
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/57772069351492564447

碩士 === 國立清華大學 === 資訊系統與應用研究所 === 96 === Music industry has gained its position of importance in the world today. Music can be created by using various types of computer software. However, users can neither conduct the playing style of music in real time. This thesis collects and analyzes the conduct data and builds a real-time conducting system so that users are able to experience the feeling of conducting a band. In this research, conducting data from ten testers are collected. These conducting data are then classified by using k-nearest neighbor rule (KNNR), Gaussian mixture model (GMM), and support vector machine (SVM) for further analysis. Finally a conducting system is constructed by using a wireless 3-dimensional acceleration sensor “Wii remote” as a conductor's baton. Beside the basic functionality of recognizing the conduct, this system can also control the volume and the speed of the music in real time.