Supervised feature learning via sparse coding for music information rerieval

O'Brien, Cian John

Title:

Supervised feature learning via sparse coding for music information rerieval

dc.contributor.advisor	Lerch, Alexander
dc.contributor.author	O'Brien, Cian John
dc.contributor.committeeMember	Freeman, Jason
dc.contributor.committeeMember	Tzanetakis, George
dc.contributor.department	Music
dc.date.accessioned	2015-06-08T18:40:22Z
dc.date.available	2015-06-08T18:40:22Z
dc.date.created	2015-05
dc.date.issued	2015-04-24
dc.date.submitted	May 2015
dc.date.updated	2015-06-08T18:40:22Z
dc.description.abstract	This thesis explores the ideas of feature learning and sparse coding for Music Information Retrieval (MIR). Sparse coding is an algorithm which aims to learn new feature representations from data automatically. In contrast to previous work which uses sparse coding in an MIR context the concept of supervised sparse coding is also investigated, which makes use of the ground-truth labels explicitly during the learning process. Here sparse coding and supervised coding are applied to two MIR problems: classification of musical genre and recognition of the emotional content of music. A variation of Label Consistent K-SVD is used to add supervision during the dictionary learning process. In the case of Music Genre Recognition (MGR) an additional discriminative term is added to encourage tracks from the same genre to have similar sparse codes. For Music Emotion Recognition (MER) a linear regression term is added to learn an optimal classifier and dictionary pair. These results indicate that while sparse coding performs well for MGR, the additional supervision fails to improve the performance. In the case of MER, supervised coding significantly outperforms both standard sparse coding and commonly used designed features, namely MFCC and pitch chroma.
dc.description.degree	M.S.
dc.format.mimetype	application/pdf
dc.identifier.uri	http://hdl.handle.net/1853/53615
dc.language.iso	en_US
dc.publisher	Georgia Institute of Technology
dc.subject	Music
dc.subject	Sparse coding
dc.subject	Music information retrieval
dc.subject	Music genre recognition
dc.subject	Music emotion recognition
dc.title	Supervised feature learning via sparse coding for music information rerieval
dc.type	Text
dc.type.genre	Thesis
dspace.entity.type	Publication
local.contributor.corporatename	College of Design
local.contributor.corporatename	School of Music
local.relation.ispartofseries	Master of Science in Music Technology
relation.isOrgUnitOfPublication	c997b6a0-7e87-4a6f-b6fc-932d776ba8d0
relation.isOrgUnitOfPublication	92d2daaa-80f2-4d99-b464-ab7c1125fc55
relation.isSeriesOfPublication	bb52c603-2646-4dfa-a9b7-9f81b43c419a
thesis.degree.level	Masters