Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks

Chi Sheng Wu; Lei Pan; Von Wun Soo

doi:10.1109/ICICE.2017.8479263

Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks

Chi Sheng Wu
, Lei Pan
, Von Wun Soo

National Tsing Hua University

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

Abstract

In this paper, we show a deep secondary learning method using convolutional neural networks (CNN) that takes into account the prediction results of a previous tagging method. In particular, we initialize the values of the kernel functions by sampling from some existing instrument signal patterns. We evaluate the tagging on male, female and no-vocal events in 100 popular songs. The performance increases 14.8%, 24.96%, 2.97% in precision, recall and accuracy respectively in comparison with the previous method.

Original language	English
Title of host publication	Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering
Subtitle of host publication	Information and Innovation for Modern Technology, ICICE 2017
Editors	Artde Donald Kin-Tak Lam, Stephen D. Prior, Teen-Hang Meen
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	412-415
Number of pages	4
ISBN (Electronic)	9781538632024
DOIs	https://doi.org/10.1109/ICICE.2017.8479263
State	Published - 01 10 2018
Externally published	Yes
Event	2017 IEEE International Conference on Information, Communication and Engineering, ICICE 2017 - Xiamen, Fujian, China Duration: 17 11 2017 → 20 11 2017

Publication series

Name	Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017

Conference

Conference	2017 IEEE International Conference on Information, Communication and Engineering, ICICE 2017
Country/Territory	China
City	Xiamen, Fujian
Period	17/11/17 → 20/11/17

Bibliographical note

Publisher Copyright:
© 2017 IEEE.

Keywords

Convolutional neural networks
Kernel initialization
Music event tagging
Secondary deep learning

Access to Document

10.1109/ICICE.2017.8479263

Cite this

Wu, C. S., Pan, L., & Soo, V. W. (2018). Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks. In A. D. K.-T. Lam, S. D. Prior, & T.-H. Meen (Eds.), Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017 (pp. 412-415). Article 8479263 (Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICICE.2017.8479263

Wu, Chi Sheng ; Pan, Lei ; Soo, Von Wun. / Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks. Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017. editor / Artde Donald Kin-Tak Lam ; Stephen D. Prior ; Teen-Hang Meen. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 412-415 (Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017).

@inproceedings{e8bae138711948c7b48dc8ee5baf57aa,

title = "Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks",

abstract = "In this paper, we show a deep secondary learning method using convolutional neural networks (CNN) that takes into account the prediction results of a previous tagging method. In particular, we initialize the values of the kernel functions by sampling from some existing instrument signal patterns. We evaluate the tagging on male, female and no-vocal events in 100 popular songs. The performance increases 14.8\%, 24.96\%, 2.97\% in precision, recall and accuracy respectively in comparison with the previous method.",

keywords = "Convolutional neural networks, Kernel initialization, Music event tagging, Secondary deep learning",

author = "Wu, \{Chi Sheng\} and Lei Pan and Soo, \{Von Wun\}",

note = "Publisher Copyright: {\textcopyright} 2017 IEEE.; 2017 IEEE International Conference on Information, Communication and Engineering, ICICE 2017 ; Conference date: 17-11-2017 Through 20-11-2017",

year = "2018",

month = oct,

day = "1",

doi = "10.1109/ICICE.2017.8479263",

language = "英语",

series = "Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "412--415",

editor = "Lam, \{Artde Donald Kin-Tak\} and Prior, \{Stephen D.\} and Teen-Hang Meen",

booktitle = "Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering",

address = "美国",

}

Wu, CS, Pan, L & Soo, VW 2018, Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks. in ADK-T Lam, SD Prior & T-H Meen (eds), Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017., 8479263, Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017, Institute of Electrical and Electronics Engineers Inc., pp. 412-415, 2017 IEEE International Conference on Information, Communication and Engineering, ICICE 2017, Xiamen, Fujian, China, 17/11/17. https://doi.org/10.1109/ICICE.2017.8479263

Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks. / Wu, Chi Sheng; Pan, Lei; Soo, Von Wun.
Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017. ed. / Artde Donald Kin-Tak Lam; Stephen D. Prior; Teen-Hang Meen. Institute of Electrical and Electronics Engineers Inc., 2018. p. 412-415 8479263 (Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017).

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

TY - GEN

T1 - Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks

AU - Wu, Chi Sheng

AU - Pan, Lei

AU - Soo, Von Wun

PY - 2018/10/1

Y1 - 2018/10/1

N2 - In this paper, we show a deep secondary learning method using convolutional neural networks (CNN) that takes into account the prediction results of a previous tagging method. In particular, we initialize the values of the kernel functions by sampling from some existing instrument signal patterns. We evaluate the tagging on male, female and no-vocal events in 100 popular songs. The performance increases 14.8%, 24.96%, 2.97% in precision, recall and accuracy respectively in comparison with the previous method.

AB - In this paper, we show a deep secondary learning method using convolutional neural networks (CNN) that takes into account the prediction results of a previous tagging method. In particular, we initialize the values of the kernel functions by sampling from some existing instrument signal patterns. We evaluate the tagging on male, female and no-vocal events in 100 popular songs. The performance increases 14.8%, 24.96%, 2.97% in precision, recall and accuracy respectively in comparison with the previous method.

KW - Convolutional neural networks

KW - Kernel initialization

KW - Music event tagging

KW - Secondary deep learning

UR - https://www.scopus.com/pages/publications/85056279536

U2 - 10.1109/ICICE.2017.8479263

DO - 10.1109/ICICE.2017.8479263

M3 - 会议稿件

AN - SCOPUS:85056279536

T3 - Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017

SP - 412

EP - 415

BT - Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering

A2 - Lam, Artde Donald Kin-Tak

A2 - Prior, Stephen D.

A2 - Meen, Teen-Hang

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2017 IEEE International Conference on Information, Communication and Engineering, ICICE 2017

Y2 - 17 November 2017 through 20 November 2017

ER -

Wu CS, Pan L, Soo VW. Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks. In Lam ADKT, Prior SD, Meen TH, editors, Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017. Institute of Electrical and Electronics Engineers Inc. 2018. p. 412-415. 8479263. (Proceedings of the 2017 IEEE International Conference on Information, Communication and Engineering: Information and Innovation for Modern Technology, ICICE 2017). doi: 10.1109/ICICE.2017.8479263

Secondary learning and kernel initialization on auto-tagging of music events using convolutional neural networks

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this