Using Music and Motion Analysis to Construct 3D Animations and Visualizations

This paper presents a study into music analysis, motion analysis and the integration of music and motion to form natural human motion in a virtual environment. Motion capture data is extracted to generate a motion library; this places the digital motion model at a fixed posture. The first step in this process is to configure the motion path curve for the database and calculate the possibility that two motions were sequential through the use of a computational algorithm. Every motion is then analyzed for the next possible smooth movement to connect to, and at the same time, an interpolation method is used to create the transitions between motions to enable the digital motion models to move fluently. Lastly, a searching algorithm sifts for possible successive motions from the motion path curve according to the music tempo. It was concluded that the higher ratio of rescaling a transition, the lower the degree of natural motion