Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Autoren

  • Mariano Cabezas
  • Sarah Kentsch
  • Luca Tomhave
  • Jens Gross
  • Maximo Larry Lopez Caceres
  • Yago Diez

Externe Organisationen

  • Universität Sydney
  • Yamagata University
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Aufsatznummer3431
Seitenumfang17
FachzeitschriftRemote sensing
Jahrgang12
Ausgabenummer20
PublikationsstatusVeröffentlicht - 19 Okt. 2020

Abstract

Deep Learning (DL) has become popular due to its ease of use and accuracy, with Transfer Learning (TL) effectively reducing the number of images needed to solve environmental problems. However, this approach has some limitations which we set out to explore: Our goal is to detect the presence of an invasive blueberry species in aerial images of wetlands. This is a key problem in ecosystem protection which is also challenging in terms of DL due to the severe imbalance present in the data. Results for the ResNet50 network show a high classification accuracy while largely ignoring the blueberry class, rendering these results of limited practical interest to detect that specific class. Moreover, by using loss function weighting and data augmentation results more akin to our practical application, our goals can be obtained. Our experiments regarding TL show that ImageNet weights do not produce satisfactory results when only the final layer of the network is trained. Furthermore, only minor gains are obtained compared with random weights when the whole network is retrained. Finally, in a study of state-of-the-art DL architectures best results were obtained by the ResNeXt architecture with 93.75 True Positive Rate and 98.11 accuracy for the Blueberry class with ResNet50, Densenet, and wideResNet obtaining close results.

ASJC Scopus Sachgebiete

Zitieren

Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data. / Cabezas, Mariano; Kentsch, Sarah; Tomhave, Luca et al.
in: Remote sensing, Jahrgang 12, Nr. 20, 3431, 19.10.2020.

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Cabezas, M., Kentsch, S., Tomhave, L., Gross, J., Caceres, M. L. L., & Diez, Y. (2020). Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data. Remote sensing, 12(20), Artikel 3431. https://doi.org/10.3390/rs12203431
Cabezas M, Kentsch S, Tomhave L, Gross J, Caceres MLL, Diez Y. Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data. Remote sensing. 2020 Okt 19;12(20):3431. doi: 10.3390/rs12203431
Cabezas, Mariano ; Kentsch, Sarah ; Tomhave, Luca et al. / Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data. in: Remote sensing. 2020 ; Jahrgang 12, Nr. 20.
Download
@article{15ad6cacd2404000a1a8002cfbc8e34d,
title = "Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data",
abstract = "Deep Learning (DL) has become popular due to its ease of use and accuracy, with Transfer Learning (TL) effectively reducing the number of images needed to solve environmental problems. However, this approach has some limitations which we set out to explore: Our goal is to detect the presence of an invasive blueberry species in aerial images of wetlands. This is a key problem in ecosystem protection which is also challenging in terms of DL due to the severe imbalance present in the data. Results for the ResNet50 network show a high classification accuracy while largely ignoring the blueberry class, rendering these results of limited practical interest to detect that specific class. Moreover, by using loss function weighting and data augmentation results more akin to our practical application, our goals can be obtained. Our experiments regarding TL show that ImageNet weights do not produce satisfactory results when only the final layer of the network is trained. Furthermore, only minor gains are obtained compared with random weights when the whole network is retrained. Finally, in a study of state-of-the-art DL architectures best results were obtained by the ResNeXt architecture with 93.75 True Positive Rate and 98.11 accuracy for the Blueberry class with ResNet50, Densenet, and wideResNet obtaining close results.",
keywords = "Data analysis, Deep learning, Transfer learning, Unbalanced data, Unmanned aerial vehicles (UAV)-acquired images",
author = "Mariano Cabezas and Sarah Kentsch and Luca Tomhave and Jens Gross and Caceres, {Maximo Larry Lopez} and Yago Diez",
year = "2020",
month = oct,
day = "19",
doi = "10.3390/rs12203431",
language = "English",
volume = "12",
journal = "Remote sensing",
issn = "2072-4292",
publisher = "Multidisciplinary Digital Publishing Institute",
number = "20",

}

Download

TY - JOUR

T1 - Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data

AU - Cabezas, Mariano

AU - Kentsch, Sarah

AU - Tomhave, Luca

AU - Gross, Jens

AU - Caceres, Maximo Larry Lopez

AU - Diez, Yago

PY - 2020/10/19

Y1 - 2020/10/19

N2 - Deep Learning (DL) has become popular due to its ease of use and accuracy, with Transfer Learning (TL) effectively reducing the number of images needed to solve environmental problems. However, this approach has some limitations which we set out to explore: Our goal is to detect the presence of an invasive blueberry species in aerial images of wetlands. This is a key problem in ecosystem protection which is also challenging in terms of DL due to the severe imbalance present in the data. Results for the ResNet50 network show a high classification accuracy while largely ignoring the blueberry class, rendering these results of limited practical interest to detect that specific class. Moreover, by using loss function weighting and data augmentation results more akin to our practical application, our goals can be obtained. Our experiments regarding TL show that ImageNet weights do not produce satisfactory results when only the final layer of the network is trained. Furthermore, only minor gains are obtained compared with random weights when the whole network is retrained. Finally, in a study of state-of-the-art DL architectures best results were obtained by the ResNeXt architecture with 93.75 True Positive Rate and 98.11 accuracy for the Blueberry class with ResNet50, Densenet, and wideResNet obtaining close results.

AB - Deep Learning (DL) has become popular due to its ease of use and accuracy, with Transfer Learning (TL) effectively reducing the number of images needed to solve environmental problems. However, this approach has some limitations which we set out to explore: Our goal is to detect the presence of an invasive blueberry species in aerial images of wetlands. This is a key problem in ecosystem protection which is also challenging in terms of DL due to the severe imbalance present in the data. Results for the ResNet50 network show a high classification accuracy while largely ignoring the blueberry class, rendering these results of limited practical interest to detect that specific class. Moreover, by using loss function weighting and data augmentation results more akin to our practical application, our goals can be obtained. Our experiments regarding TL show that ImageNet weights do not produce satisfactory results when only the final layer of the network is trained. Furthermore, only minor gains are obtained compared with random weights when the whole network is retrained. Finally, in a study of state-of-the-art DL architectures best results were obtained by the ResNeXt architecture with 93.75 True Positive Rate and 98.11 accuracy for the Blueberry class with ResNet50, Densenet, and wideResNet obtaining close results.

KW - Data analysis

KW - Deep learning

KW - Transfer learning

KW - Unbalanced data

KW - Unmanned aerial vehicles (UAV)-acquired images

UR - http://www.scopus.com/inward/record.url?scp=85092778376&partnerID=8YFLogxK

U2 - 10.3390/rs12203431

DO - 10.3390/rs12203431

M3 - Article

VL - 12

JO - Remote sensing

JF - Remote sensing

SN - 2072-4292

IS - 20

M1 - 3431

ER -

Von denselben Autoren