Democritus University of Thrace (DUTH)
Audio source separation with machine learning techniques
Abstract
dc:descriptionThis thesis addresses the problem of Blind Source Separation (BSS) and possible solutions in some of its forms. It focuses on a series of different scenarios, where various implementations tackle the BSS problem. The frameworks presented in the thesis are separated in two categories, depending on the scientific approach of their implementation. The first category includes methods that employed statistical signal processing, while the second includes deep learning techniques. All the implementations are mathematically supported and successfully evaluated after extended experimentation.
Degree
thesis:*- Grantor dc:publisher
- Democritus University of Thrace (DUTH)
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Sgouros, Thomas
- Σγούρος, Θωμάς
Subjects
dc:subject × 18- Τυφλός διαχωρισμός πηγών
- Επεξεργασία ήχου
- Μηχανική μάθηση
- Βαθιά μάθηση
- Ψηφιακή επεξεργασία σημάτων
- Αναγνώριση προτύπων
- Blind source separation
- Sound processing
- Machine learning
- Deep learning
- Digital signal processing
- Pattern recognition
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Υπολογιστές, Υλικό (hardware) και Αρχιτεκτονική
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Computer science, Hardware and Architecture
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/53199