Music Genre Classification using Masked Conditional Neural Networks

Fady Medhat; Edwin David Chesmore; John Allen Robinson

doi:10.1007/978-3-319-70096-0_49

Music Genre Classification using Masked Conditional Neural Networks

Fady Medhat, Edwin David Chesmore, John Allen Robinson

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

The ConditionaL Neural Networks (CLNN) and the Masked ConditionaL Neural Networks (MCLNN) exploit the nature of multi-dimensional temporal signals. The CLNN captures the conditional temporal influence between the frames in a window and the mask in the MCLNN enforces a systematic sparseness that follows a filterbank-like pattern over the network links. The mask induces the network to learn about time-frequency representations in bands, allowing the network to sustain frequency shifts. Additionally, the mask in the MCLNN automates the exploration of a range of feature combinations, usually done through an exhaustive manual search. We have evaluated the MCLNN performance using the Ballroom and Homburg datasets of music genres. MCLNN has achieved accuracies that are competitive to state-of-the-art handcrafted attempts in addition to models based on Convolutional Neural Networks.

Original language	English
Title of host publication	Neural Information Processing
Subtitle of host publication	24th International Conference, ICONIP 2017, Guangzhou, China, November 14-18, 2017, Proceedings, Part II
Publisher	Springer
ISBN (Print)	978-3-319-70095-3
DOIs	https://doi.org/10.1007/978-3-319-70096-0_49
Publication status	Published - 18 Feb 2018
Event	International Conference on Neural Information Processing - Guangzhou, China Duration: 14 Nov 2017 → 18 Nov 2017

Publication series

Name	Lecture Notes in Computer Science

Conference

Conference	International Conference on Neural Information Processing
Abbreviated title	ICONIP
Country/Territory	China
City	Guangzhou
Period	14/11/17 → 18/11/17

Keywords

MCLNN
CLNN
RBM

Access to Document

10.1007/978-3-319-70096-0_49

1802.06432v1Accepted author manuscript, 1.19 MB

https://arxiv.org/abs/1802.06432

Cite this

@inproceedings{359a80adbe074cec9b8cf145ca7a8496,

title = "Music Genre Classification using Masked Conditional Neural Networks",

abstract = " The ConditionaL Neural Networks (CLNN) and the Masked ConditionaL Neural Networks (MCLNN) exploit the nature of multi-dimensional temporal signals. The CLNN captures the conditional temporal influence between the frames in a window and the mask in the MCLNN enforces a systematic sparseness that follows a filterbank-like pattern over the network links. The mask induces the network to learn about time-frequency representations in bands, allowing the network to sustain frequency shifts. Additionally, the mask in the MCLNN automates the exploration of a range of feature combinations, usually done through an exhaustive manual search. We have evaluated the MCLNN performance using the Ballroom and Homburg datasets of music genres. MCLNN has achieved accuracies that are competitive to state-of-the-art handcrafted attempts in addition to models based on Convolutional Neural Networks. ",

keywords = "MCLNN, CLNN, RBM",

author = "Fady Medhat and Chesmore, {Edwin David} and Robinson, {John Allen}",

year = "2018",

month = feb,

day = "18",

doi = "10.1007/978-3-319-70096-0_49",

language = "English",

isbn = "978-3-319-70095-3",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

booktitle = "Neural Information Processing",

address = "Germany",

note = "International Conference on Neural Information Processing, ICONIP ; Conference date: 14-11-2017 Through 18-11-2017",

}

Medhat, F , Chesmore, ED & Robinson, JA 2018, Music Genre Classification using Masked Conditional Neural Networks. in Neural Information Processing: 24th International Conference, ICONIP 2017, Guangzhou, China, November 14-18, 2017, Proceedings, Part II. Lecture Notes in Computer Science, Springer, International Conference on Neural Information Processing, Guangzhou, China, 14/11/17. https://doi.org/10.1007/978-3-319-70096-0_49

Music Genre Classification using Masked Conditional Neural Networks. / Medhat, Fady ; Chesmore, Edwin David ; Robinson, John Allen.
Neural Information Processing: 24th International Conference, ICONIP 2017, Guangzhou, China, November 14-18, 2017, Proceedings, Part II. Springer, 2018. (Lecture Notes in Computer Science).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Music Genre Classification using Masked Conditional Neural Networks

AU - Medhat, Fady

AU - Chesmore, Edwin David

AU - Robinson, John Allen

PY - 2018/2/18

Y1 - 2018/2/18

N2 - The ConditionaL Neural Networks (CLNN) and the Masked ConditionaL Neural Networks (MCLNN) exploit the nature of multi-dimensional temporal signals. The CLNN captures the conditional temporal influence between the frames in a window and the mask in the MCLNN enforces a systematic sparseness that follows a filterbank-like pattern over the network links. The mask induces the network to learn about time-frequency representations in bands, allowing the network to sustain frequency shifts. Additionally, the mask in the MCLNN automates the exploration of a range of feature combinations, usually done through an exhaustive manual search. We have evaluated the MCLNN performance using the Ballroom and Homburg datasets of music genres. MCLNN has achieved accuracies that are competitive to state-of-the-art handcrafted attempts in addition to models based on Convolutional Neural Networks.

AB - The ConditionaL Neural Networks (CLNN) and the Masked ConditionaL Neural Networks (MCLNN) exploit the nature of multi-dimensional temporal signals. The CLNN captures the conditional temporal influence between the frames in a window and the mask in the MCLNN enforces a systematic sparseness that follows a filterbank-like pattern over the network links. The mask induces the network to learn about time-frequency representations in bands, allowing the network to sustain frequency shifts. Additionally, the mask in the MCLNN automates the exploration of a range of feature combinations, usually done through an exhaustive manual search. We have evaluated the MCLNN performance using the Ballroom and Homburg datasets of music genres. MCLNN has achieved accuracies that are competitive to state-of-the-art handcrafted attempts in addition to models based on Convolutional Neural Networks.

KW - MCLNN

KW - CLNN

KW - RBM

U2 - 10.1007/978-3-319-70096-0_49

DO - 10.1007/978-3-319-70096-0_49

M3 - Conference contribution

SN - 978-3-319-70095-3

T3 - Lecture Notes in Computer Science

BT - Neural Information Processing

PB - Springer

T2 - International Conference on Neural Information Processing

Y2 - 14 November 2017 through 18 November 2017

ER -