Improving Upon Musical Analyses Of Conducting Gestures Using Computer Vision.
Teresa Marrin Nakra, Daniel Steven Tilden, Andrea Selinger Salgian · University of Michigan Library Repository · 2010
For more than ten years, researchers have been developing software-based methods for analyzing the movements of orchestral conductors.Beginning with the Conductor's Jacket research project, researchers have combined wearable sensors with face-on video recordings to improve tracking and understanding of the structure inherent in the gestures.Building upon more recent work employing computer vision, the current project refines the methods for tracking the hands and visualizes the data in a way that reveals more of the underlying structure in an easier-toview fashion.Such improved methods will enable more specific understanding of the functions of conducting gestures, and allow for more concrete applications of those gestures to interactive conducting systems for public exhibits, video games, and integrating dynamic audio enhancements to live concerts.