The Group ‘Perception, Action, and Interaction’ at NICS explores the interaction between humans and digital media; ‘Perception’ stands for audio analysis, psychoacoustics and visual cognition; ‘Action’ stands for designing sound and visual arts; ‘Interaction’ stands for how humans interfaces with the digital environment. One ongoing research example is the sonification of human body movement with application purposes in music therapy.

One of the devices used by the laboratory to sense the body movements is the Microsoft Kinect. Therefore, some projects rely on a robust tool to preprocess the incoming data and extract body movement features. My role is to develop that tool and make it available as an user friendly software for people with low level of knowledge in software development.

The main features implemented in the library available on my github are:

• Interface with KinectV2
• Smooth skeleton
• Calibrate for the floor position and room size
• Extract body features
• Send features through network via OSC

The video below is a brief demonstration of the library and its features.