Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification

Yordan Yasin; Dewinda Julianensi Rumala; Mauridhi Hery Purnomo; Anak Agung Putri Ratna; Afif Nurul Hidayati; Ingrid Nurtanio; Reza Fuad Rachmadi; I. Ketut Eddy Purnama

doi:10.1109/CENIM51130.2020.9297994

Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification

Yordan Yasin, Dewinda Julianensi Rumala, Mauridhi Hery Purnomo, Anak Agung Putri Ratna, Afif Nurul Hidayati, Ingrid Nurtanio, Reza Fuad Rachmadi, I. Ketut Eddy Purnama

Department of Electrical Engineering

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

1 Citation (Scopus)

Abstract

A computerized skin disease classification system generally works on closed-set data, meaning images from unknown classes will still be classified as one of the known classes. In the Teledermatology system, skin disease classes are usually defined before the training process. However, in the real world application, it may receive images that belong to a new class or disease. To avoid misclassification, we have implemented the Extreme Value Theory of Weibull distribution function for out of distribution detection and incorporated the OpenMax layer to the deep networks for open-set recognition in skin disease classification. The system can classify seven classes of common skin disease in Indonesia with an accuracy of 71.64% using Inception v3 for closed-set data, while it achieved an accuracy of 83.33% for open-set recognition. The result indicates that the proposed method in this study has reached the purpose of recognizing open-set data in skin disease classification.

Original language	English
Title of host publication	CENIM 2020 - Proceeding
Subtitle of host publication	International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	332-337
Number of pages	6
ISBN (Electronic)	9781728182834
DOIs	https://doi.org/10.1109/CENIM51130.2020.9297994
Publication status	Published - 17 Nov 2020
Event	2020 International Conference on Computer Engineering, Network, and Intelligent Multimedia, CENIM 2020 - Virtual, Surabaya, Indonesia Duration: 17 Nov 2020 → 18 Nov 2020

Publication series

Name	CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020

Conference

Conference	2020 International Conference on Computer Engineering, Network, and Intelligent Multimedia, CENIM 2020
Country/Territory	Indonesia
City	Virtual, Surabaya
Period	17/11/20 → 18/11/20

Keywords

Deep Learning
Extreme Value Theory
Image Classification
Open Set Recognition
Skin Disease

Access to Document

10.1109/CENIM51130.2020.9297994

Cite this

Yasin, Y., Rumala, D. J., Purnomo, M. H., Ratna, A. A. P., Hidayati, A. N., Nurtanio, I., Rachmadi, R. F., & Purnama, I. K. E. (2020). Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification. In CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020 (pp. 332-337). Article 9297994 (CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CENIM51130.2020.9297994

Yasin, Yordan ; Rumala, Dewinda Julianensi ; Purnomo, Mauridhi Hery et al. / Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification. CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020. Institute of Electrical and Electronics Engineers Inc., 2020. pp. 332-337 (CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020).

@inproceedings{2478199d2835499bbee18cfbc42998f3,

title = "Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification",

abstract = "A computerized skin disease classification system generally works on closed-set data, meaning images from unknown classes will still be classified as one of the known classes. In the Teledermatology system, skin disease classes are usually defined before the training process. However, in the real world application, it may receive images that belong to a new class or disease. To avoid misclassification, we have implemented the Extreme Value Theory of Weibull distribution function for out of distribution detection and incorporated the OpenMax layer to the deep networks for open-set recognition in skin disease classification. The system can classify seven classes of common skin disease in Indonesia with an accuracy of 71.64% using Inception v3 for closed-set data, while it achieved an accuracy of 83.33% for open-set recognition. The result indicates that the proposed method in this study has reached the purpose of recognizing open-set data in skin disease classification.",

keywords = "Deep Learning, Extreme Value Theory, Image Classification, Open Set Recognition, Skin Disease",

author = "Yordan Yasin and Rumala, {Dewinda Julianensi} and Purnomo, {Mauridhi Hery} and Ratna, {Anak Agung Putri} and Hidayati, {Afif Nurul} and Ingrid Nurtanio and Rachmadi, {Reza Fuad} and Purnama, {I. Ketut Eddy}",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE. Copyright: Copyright 2021 Elsevier B.V., All rights reserved.; 2020 International Conference on Computer Engineering, Network, and Intelligent Multimedia, CENIM 2020 ; Conference date: 17-11-2020 Through 18-11-2020",

year = "2020",

month = nov,

day = "17",

doi = "10.1109/CENIM51130.2020.9297994",

language = "English",

series = "CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020",

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

pages = "332--337",

booktitle = "CENIM 2020 - Proceeding",

address = "United States",

}

Yasin, Y, Rumala, DJ, Purnomo, MH, Ratna, AAP, Hidayati, AN, Nurtanio, I, Rachmadi, RF & Purnama, IKE 2020, Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification. in CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020., 9297994, CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020, Institute of Electrical and Electronics Engineers Inc., pp. 332-337, 2020 International Conference on Computer Engineering, Network, and Intelligent Multimedia, CENIM 2020, Virtual, Surabaya, Indonesia, 17/11/20. https://doi.org/10.1109/CENIM51130.2020.9297994

Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification. / Yasin, Yordan; Rumala, Dewinda Julianensi; Purnomo, Mauridhi Hery et al.
CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020. Institute of Electrical and Electronics Engineers Inc., 2020. p. 332-337 9297994 (CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020).

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

TY - GEN

T1 - Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification

AU - Yasin, Yordan

AU - Rumala, Dewinda Julianensi

AU - Purnomo, Mauridhi Hery

AU - Ratna, Anak Agung Putri

AU - Hidayati, Afif Nurul

AU - Nurtanio, Ingrid

AU - Rachmadi, Reza Fuad

AU - Purnama, I. Ketut Eddy

PY - 2020/11/17

Y1 - 2020/11/17

N2 - A computerized skin disease classification system generally works on closed-set data, meaning images from unknown classes will still be classified as one of the known classes. In the Teledermatology system, skin disease classes are usually defined before the training process. However, in the real world application, it may receive images that belong to a new class or disease. To avoid misclassification, we have implemented the Extreme Value Theory of Weibull distribution function for out of distribution detection and incorporated the OpenMax layer to the deep networks for open-set recognition in skin disease classification. The system can classify seven classes of common skin disease in Indonesia with an accuracy of 71.64% using Inception v3 for closed-set data, while it achieved an accuracy of 83.33% for open-set recognition. The result indicates that the proposed method in this study has reached the purpose of recognizing open-set data in skin disease classification.

AB - A computerized skin disease classification system generally works on closed-set data, meaning images from unknown classes will still be classified as one of the known classes. In the Teledermatology system, skin disease classes are usually defined before the training process. However, in the real world application, it may receive images that belong to a new class or disease. To avoid misclassification, we have implemented the Extreme Value Theory of Weibull distribution function for out of distribution detection and incorporated the OpenMax layer to the deep networks for open-set recognition in skin disease classification. The system can classify seven classes of common skin disease in Indonesia with an accuracy of 71.64% using Inception v3 for closed-set data, while it achieved an accuracy of 83.33% for open-set recognition. The result indicates that the proposed method in this study has reached the purpose of recognizing open-set data in skin disease classification.

KW - Deep Learning

KW - Extreme Value Theory

KW - Image Classification

KW - Open Set Recognition

KW - Skin Disease

UR - http://www.scopus.com/inward/record.url?scp=85099647969&partnerID=8YFLogxK

U2 - 10.1109/CENIM51130.2020.9297994

DO - 10.1109/CENIM51130.2020.9297994

M3 - Conference contribution

AN - SCOPUS:85099647969

T3 - CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020

SP - 332

EP - 337

BT - CENIM 2020 - Proceeding

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2020 International Conference on Computer Engineering, Network, and Intelligent Multimedia, CENIM 2020

Y2 - 17 November 2020 through 18 November 2020

ER -

Yasin Y, Rumala DJ, Purnomo MH, Ratna AAP, Hidayati AN, Nurtanio I et al. Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification. In CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020. Institute of Electrical and Electronics Engineers Inc. 2020. p. 332-337. 9297994. (CENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020). doi: 10.1109/CENIM51130.2020.9297994

Open Set Deep Networks Based on Extreme Value Theory (EVT) for Open Set Recognition in Skin Disease Classification

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this