ABSTRACT
We propose a process model for hierarchical perceptual sound organization that enables the recognition of perceptual sounds in incoming sound signals. We consider perceptual sound organization as a scene analysis problem in the auditory domain. Our current application is a music scene analysis system that recognizes rhythm, chords, and source-separated musical notes included in incoming music signals. The process model consists of multiple processing modules and a probability network for information integration. Its structure is conceptually based on the blackboard architecture. However, employment of the Bayesian probability network has facilitated integration of multiple sources of information provided by autonomous modules without global control knowledge.
