Sound event detection and extraction for automatic music transcription
Metadata[+] Show full item record
[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] This dissertation addresses the problem of automatic music transcription. Transcription is an open problem in the area of musical analysis and audio signal processing, and one that has drawn increased attention in recent years. Despite the increase in related research, a system that is capable of transcription in the general sense does not appear to be on the horizon. In this document, we propose a new approach to the problem, motivated by the human approach to transcription, that is hoped to be a foundation for constructing a viable system in the future. Additionally, we investigate a method of onset and offset detection, based upon a model of human auditory physiology, for robustly identifying temporal boundaries of sound events while retaining information of benefit to subsequent stages of processing within the transcription framework.
Access is limited to the campus of the University of Missouri--Columbia.