National and Kapodistrian University of Athens
Επεξεργασία μουσικών σημάτων για μουσικολογικές εφαρμογές
Abstract
dc:descriptionIn this thesis the development of content-based techniques for the extraction and study of periodicities and rhythmic characteristics of music recordings took place. This thesis focused on the study of traditional music genres that haven t been studied in the literature. The methodology exhibited refers to the extraction of rhythm and music meter of a recording the extraction of a music summary the finding of similar rhythmic recordings and the modeling and tracking of rhythmic patterns. To this a feature extraction step takes place and energy and spectral features are extracted. For the application of periodicities and summary extraction the Self Similarity Matrix is employed. For the finding of similar rhythmic recordings the mean diagonal of the above matrix of each recording is compared with the diagonals of other recordings using the Dynamic Time Warping technique. Finally, for the modeling and tracking of rhythmic patterns the Hidden Markov Models are employed and each HMM state represents a rhythmic pattern state. Whenever an occurrence of the HMM states is found then a rhythmic pattern is considered to be located.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Antonopoulos, Iasonas
- Αντωνόπουλος, Ιάσονας
Subjects
dc:subject × 14- Επεξεργασία μουσικών σημάτων
- Αναγνώριση προτύπων
- Εξαγωγή περιοδικοτήτων
- Δυναμική στρέβλωση στο πεδίο του χρόνου
- Μαρκοβιανά μοντέλα
- Music signal processing
- Pattern recognition
- Periodicity extraction
- Dynamic time warping
- Hidden Markov models
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Natural Sciences
- Computer and Information Sciences
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/23951