Identification of Saimaa Ringed Seal Individuals Using Transfer Learning

E. Nepovinnykh; T. Eerola; H. Kälviäinen; G. Radchenko

doi:10.1007/978-3-030-01449-0_18

Identification of Saimaa Ringed Seal Individuals Using Transfer Learning

E. Nepovinnykh, T. Eerola, H. Kälviäinen, G. Radchenko

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

Abstract

The conservation efforts of the endangered Saimaa ringed seal depend on the ability to reliably estimate the population size and to track individuals. Wildlife photo-identification has been successfully utilized in monitoring for various species. Traditionally, the collected images have been analyzed by biologists. However, due to the rapid increase in the amount of image data, there is a demand for automated methods. Ringed seals have pelage patterns that are unique to each seal enabling the individual identification. In this work, two methods of Saimaa ringed seal identification based on transfer learning are proposed. The first method involves retraining of an existing convolutional neural network (CNN). The second method uses the CNN trained for image classification to extract features which are then used to train a Support Vector Machine (SVM) classifier. Both approaches show over 90% identification accuracy on challenging image data, the SVM based method being slightly better. © 2018, Springer Nature Switzerland AG.

Original language	English
Title of host publication	International Conference on Advanced Concepts for Intelligent Vision Systems
Subtitle of host publication	ACIVS 2018: Advanced Concepts for Intelligent Vision Systems
Pages	211-222
Number of pages	12
Volume	11182 LNCS
DOIs	https://doi.org/10.1007/978-3-030-01449-0_18
Publication status	Published - 2018
Event	19th International Conference on Advanced Concepts for Intelligent Vision Systems - Poitiers, France Duration: 20 Sept 2018 → 27 Sept 2018 http://acivs.org/acivs2018/

Conference

Conference	19th International Conference on Advanced Concepts for Intelligent Vision Systems
Abbreviated title	ACIVS 2018
Country/Territory	France
City	Poitiers
Period	20/09/18 → 27/09/18
Internet address	http://acivs.org/acivs2018/

Keywords

Animal biometrics
Convolutional neural networks
Identification
Image segmentation
Saimaa ringed seals
Transfer learning
Animals
Computer vision
Convolution
Identification (control systems)
Image retrieval
Neural networks
Population statistics
Support vector machines
Automated methods
Convolutional neural network
Convolutional Neural Networks (CNN)
Identification accuracy
Individual identification
Photo identification
Ringed seals

Access to Document

10.1007/978-3-030-01449-0_18

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85054839175&doi=10.1007%2f978-3-030-01449-0_18&partnerID=40&md5=d1c5b3d61a98e7623a9db6f4a0a6034e

Cite this

@inproceedings{38f35d824d254c2d992858dcf235c7e0,

title = "Identification of Saimaa Ringed Seal Individuals Using Transfer Learning",

abstract = "The conservation efforts of the endangered Saimaa ringed seal depend on the ability to reliably estimate the population size and to track individuals. Wildlife photo-identification has been successfully utilized in monitoring for various species. Traditionally, the collected images have been analyzed by biologists. However, due to the rapid increase in the amount of image data, there is a demand for automated methods. Ringed seals have pelage patterns that are unique to each seal enabling the individual identification. In this work, two methods of Saimaa ringed seal identification based on transfer learning are proposed. The first method involves retraining of an existing convolutional neural network (CNN). The second method uses the CNN trained for image classification to extract features which are then used to train a Support Vector Machine (SVM) classifier. Both approaches show over 90% identification accuracy on challenging image data, the SVM based method being slightly better. {\textcopyright} 2018, Springer Nature Switzerland AG.",

keywords = "Animal biometrics, Convolutional neural networks, Identification, Image segmentation, Saimaa ringed seals, Transfer learning, Animals, Computer vision, Convolution, Identification (control systems), Image retrieval, Neural networks, Population statistics, Support vector machines, Automated methods, Convolutional neural network, Convolutional Neural Networks (CNN), Identification accuracy, Individual identification, Photo identification, Ringed seals",

author = "E. Nepovinnykh and T. Eerola and H. K{\"a}lvi{\"a}inen and G. Radchenko",

note = "Conference code: 218999 Cited By :2 Export Date: 27 August 2021 Correspondence Address: Eerola, T.; Machine Vision and Pattern Recognition Laboratory, Finland; email: tuomas.eerola@lut.fi; 19th International Conference on Advanced Concepts for Intelligent Vision Systems, ACIVS 2018 ; Conference date: 20-09-2018 Through 27-09-2018",

year = "2018",

doi = "10.1007/978-3-030-01449-0_18",

language = "English",

isbn = "03029743 (ISSN); 9783030014483 (ISBN)",

volume = "11182 LNCS",

pages = "211--222",

booktitle = "International Conference on Advanced Concepts for Intelligent Vision Systems",

url = "http://acivs.org/acivs2018/",

}

Nepovinnykh, E, Eerola, T, Kälviäinen, H & Radchenko, G 2018, Identification of Saimaa Ringed Seal Individuals Using Transfer Learning. in International Conference on Advanced Concepts for Intelligent Vision Systems: ACIVS 2018: Advanced Concepts for Intelligent Vision Systems . vol. 11182 LNCS, pp. 211-222, 19th International Conference on Advanced Concepts for Intelligent Vision Systems, Poitiers, France, 20/09/18. https://doi.org/10.1007/978-3-030-01449-0_18

Identification of Saimaa Ringed Seal Individuals Using Transfer Learning. / Nepovinnykh, E.; Eerola, T.; Kälviäinen, H. et al.
International Conference on Advanced Concepts for Intelligent Vision Systems: ACIVS 2018: Advanced Concepts for Intelligent Vision Systems . Vol. 11182 LNCS 2018. p. 211-222.

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

TY - GEN

T1 - Identification of Saimaa Ringed Seal Individuals Using Transfer Learning

AU - Nepovinnykh, E.

AU - Eerola, T.

AU - Kälviäinen, H.

AU - Radchenko, G.

N1 - Conference code: 218999 Cited By :2 Export Date: 27 August 2021 Correspondence Address: Eerola, T.; Machine Vision and Pattern Recognition Laboratory, Finland; email: tuomas.eerola@lut.fi

PY - 2018

Y1 - 2018

N2 - The conservation efforts of the endangered Saimaa ringed seal depend on the ability to reliably estimate the population size and to track individuals. Wildlife photo-identification has been successfully utilized in monitoring for various species. Traditionally, the collected images have been analyzed by biologists. However, due to the rapid increase in the amount of image data, there is a demand for automated methods. Ringed seals have pelage patterns that are unique to each seal enabling the individual identification. In this work, two methods of Saimaa ringed seal identification based on transfer learning are proposed. The first method involves retraining of an existing convolutional neural network (CNN). The second method uses the CNN trained for image classification to extract features which are then used to train a Support Vector Machine (SVM) classifier. Both approaches show over 90% identification accuracy on challenging image data, the SVM based method being slightly better. © 2018, Springer Nature Switzerland AG.

AB - The conservation efforts of the endangered Saimaa ringed seal depend on the ability to reliably estimate the population size and to track individuals. Wildlife photo-identification has been successfully utilized in monitoring for various species. Traditionally, the collected images have been analyzed by biologists. However, due to the rapid increase in the amount of image data, there is a demand for automated methods. Ringed seals have pelage patterns that are unique to each seal enabling the individual identification. In this work, two methods of Saimaa ringed seal identification based on transfer learning are proposed. The first method involves retraining of an existing convolutional neural network (CNN). The second method uses the CNN trained for image classification to extract features which are then used to train a Support Vector Machine (SVM) classifier. Both approaches show over 90% identification accuracy on challenging image data, the SVM based method being slightly better. © 2018, Springer Nature Switzerland AG.

KW - Animal biometrics

KW - Convolutional neural networks

KW - Identification

KW - Image segmentation

KW - Saimaa ringed seals

KW - Transfer learning

KW - Animals

KW - Computer vision

KW - Convolution

KW - Identification (control systems)

KW - Image retrieval

KW - Neural networks

KW - Population statistics

KW - Support vector machines

KW - Automated methods

KW - Convolutional neural network

KW - Convolutional Neural Networks (CNN)

KW - Identification accuracy

KW - Individual identification

KW - Photo identification

KW - Ringed seals

U2 - 10.1007/978-3-030-01449-0_18

DO - 10.1007/978-3-030-01449-0_18

M3 - Conference Paper

SN - 03029743 (ISSN); 9783030014483 (ISBN)

VL - 11182 LNCS

SP - 211

EP - 222

BT - International Conference on Advanced Concepts for Intelligent Vision Systems

T2 - 19th International Conference on Advanced Concepts for Intelligent Vision Systems

Y2 - 20 September 2018 through 27 September 2018

ER -

Identification of Saimaa Ringed Seal Individuals Using Transfer Learning

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this