Hartmann, Martin (JY): Modelling Human Perception and Cognition of Music Segmentation
While listening to a piece, we naturally divide it into sections, such as introduction, verse and chorus. The research examines segmentation from the perspective of its corresponding musical features and perceptual organizational principles. The computational modelling of segmentation can be greatly improved, since recent efforts did not exploit cognitive aspects and were limited to monophony or insufficient styles.
Our highly interdisciplinary approach combines music cognition, acoustics, signal processing and computer science methods. New data collection and analysis strategies are suggested in order to model segment cognition using robust features.
This work deepens our knowledge on temporal processing of perceptual streams found in music, speech and movement. Further, it has strong practical implications for the digital online music industry and sets a pathway towards understanding how music influences our brain, kinesis, and emotions.