Sonification and augmented data sets in binary classification

Fares El-Azm

AbstractIn the thesis an auditory browser based on granular synthesis is designed and implemented to aid in browsing through long EEG time courses. This application can be used when ICA is applied to EEG signals as a means of decontamination and is intended to accelerate the identification of artifactual time courses, though this was not confirmed through testing. Furthermore, an introduction to the rather young field of sonification and EEG sonification is presented, also including introductory chapters on auditory perception and sound synthesis. Concepts in classification are introduced and the idea of augmented data sets using PCA and ICA is investigated. It is shown that augmenting data sets can "supervise" PCA and ICA, though this was seen to be especially true for PCA.
KeywordsSonification, classification, EEG, granular synthesis, auditory perception, sound synthesis, augmented data sets, ICA, PCA
TypeMaster's thesis [Academic thesis]
PublisherInformatics and Mathematical Modelling, Technical University of Denmark, DTU
AddressRichard Petersens Plads, Building 321, DK-2800 Kgs. Lyngby
NoteSupervised by Prof. Lars Kai Hansen
Electronic version(s)[pdf]
BibTeX data [bibtex]
IMM Group(s)Intelligent Signal Processing