Classification of Ecological Data by Deep Learning

Shaobo Liu; Frank Y. Shih; Gareth Russell; Kimberly Russell; Hai Phan

doi:https://doi.org/10.1142/S0218001420520102

Classification of Ecological Data by Deep Learning

Shaobo Liu, Frank Y. Shih, Gareth Russell, Kimberly Russell, Hai Phan

New Jersey Institute of Technology

Research output: Contribution to journal › Article › peer-review

1 Scopus citations

Abstract

Ecologists have been studying different computational models in the classification of ecological species. In this paper, we intend to take advantages of variant deep-learning models, including LeNet, AlexNet, VGG models, residual neural network, and inception models, to classify ecological datasets, such as bee wing and butterfly. Since the datasets contain relatively small data samples and unbalanced samples in each class, we apply data augmentation and transfer learning techniques. Furthermore, newly designed inception residual and inception modules are developed to enhance feature extraction and increase classification rates. As comparing against currently available deep-learning models, experimental results show that the proposed inception residual block can avoid the vanishing gradient problem and achieve a high accuracy rate of 92%.

Original language	English (US)
Article number	2052010
Journal	International Journal of Pattern Recognition and Artificial Intelligence
Volume	34
Issue number	13
DOIs	https://doi.org/10.1142/S0218001420520102
State	Published - Dec 15 2020

ASJC Scopus subject areas

Software
Computer Vision and Pattern Recognition
Artificial Intelligence

Keywords

Deep learning
bee wings
convolutional neural network
ecology
image augmentation
image classification
inception residual module

Access to Document

https://doi.org/10.1142/S0218001420520102

Cite this

@article{54a499dd8e5b45269b10192d66ddab48,

title = "Classification of Ecological Data by Deep Learning",

abstract = "Ecologists have been studying different computational models in the classification of ecological species. In this paper, we intend to take advantages of variant deep-learning models, including LeNet, AlexNet, VGG models, residual neural network, and inception models, to classify ecological datasets, such as bee wing and butterfly. Since the datasets contain relatively small data samples and unbalanced samples in each class, we apply data augmentation and transfer learning techniques. Furthermore, newly designed inception residual and inception modules are developed to enhance feature extraction and increase classification rates. As comparing against currently available deep-learning models, experimental results show that the proposed inception residual block can avoid the vanishing gradient problem and achieve a high accuracy rate of 92%. ",

keywords = "Deep learning, bee wings, convolutional neural network, ecology, image augmentation, image classification, inception residual module",

author = "Shaobo Liu and Shih, {Frank Y.} and Gareth Russell and Kimberly Russell and Hai Phan",

note = "Funding Information: BS from the National Cheng Kung University, Tainan, Taiwan, in 1980, MS from the State Uni- versity of New York, Stony Brook, USA, in 1983, and PhD from the Purdue University, West Lafayette, Indiana, USA, in 1987. He is a Professor jointly appointed in the Department of Computer Science, the Department of Electrical and Computer Engineering, and the Department of Biomedical Engineering at New Jersey Institute of Technology, Newark, New Jersey. He currently serves as the Director of Arti¯cial Intelligence and Computer Vision Laboratory. Dr. Shih held a visiting professor position at the Princeton University, Columbia University, National Taiwan University, National Institute of Informatics, Tokyo, Conservatoire National Des Arts Et Metiers, Paris, and Nanjing University of Information Science and Technology, China. He is an internationally renowned scholar and currently serves as Editor-in-Chief for the International Journal of Pattern Recognition and Arti¯cial Intelligence. He was Editor-in-Chief for the International Journal of Multimedia Intelligence and Security. In addition, he is on the Editorial Board of 12 international journals. He has served as a steering member, session chair, and committee member for numerous professional conferences and workshops. He has received numerous grants from National Science Foundation, NIH, NASA, Navy and Air Force, and Industry. He has won the Research Initiation Award from NSF, the Outstanding Teaching Award and the Board of Overseers Excellence in Research Award from NJIT, and the Best Paper Awards from journals and conferences. Publisher Copyright: {\textcopyright} 2020 World Scientific Publishing Company.",

year = "2020",

month = dec,

day = "15",

doi = "https://doi.org/10.1142/S0218001420520102",

language = "English (US)",

volume = "34",

journal = "International Journal of Pattern Recognition and Artificial Intelligence",

issn = "0218-0014",

publisher = "World Scientific Publishing Co. Pte Ltd",

number = "13",

}

TY - JOUR

T1 - Classification of Ecological Data by Deep Learning

AU - Liu, Shaobo

AU - Shih, Frank Y.

AU - Russell, Gareth

AU - Russell, Kimberly

AU - Phan, Hai

N1 - Funding Information: BS from the National Cheng Kung University, Tainan, Taiwan, in 1980, MS from the State Uni- versity of New York, Stony Brook, USA, in 1983, and PhD from the Purdue University, West Lafayette, Indiana, USA, in 1987. He is a Professor jointly appointed in the Department of Computer Science, the Department of Electrical and Computer Engineering, and the Department of Biomedical Engineering at New Jersey Institute of Technology, Newark, New Jersey. He currently serves as the Director of Arti¯cial Intelligence and Computer Vision Laboratory. Dr. Shih held a visiting professor position at the Princeton University, Columbia University, National Taiwan University, National Institute of Informatics, Tokyo, Conservatoire National Des Arts Et Metiers, Paris, and Nanjing University of Information Science and Technology, China. He is an internationally renowned scholar and currently serves as Editor-in-Chief for the International Journal of Pattern Recognition and Arti¯cial Intelligence. He was Editor-in-Chief for the International Journal of Multimedia Intelligence and Security. In addition, he is on the Editorial Board of 12 international journals. He has served as a steering member, session chair, and committee member for numerous professional conferences and workshops. He has received numerous grants from National Science Foundation, NIH, NASA, Navy and Air Force, and Industry. He has won the Research Initiation Award from NSF, the Outstanding Teaching Award and the Board of Overseers Excellence in Research Award from NJIT, and the Best Paper Awards from journals and conferences. Publisher Copyright: © 2020 World Scientific Publishing Company.

PY - 2020/12/15

Y1 - 2020/12/15

N2 - Ecologists have been studying different computational models in the classification of ecological species. In this paper, we intend to take advantages of variant deep-learning models, including LeNet, AlexNet, VGG models, residual neural network, and inception models, to classify ecological datasets, such as bee wing and butterfly. Since the datasets contain relatively small data samples and unbalanced samples in each class, we apply data augmentation and transfer learning techniques. Furthermore, newly designed inception residual and inception modules are developed to enhance feature extraction and increase classification rates. As comparing against currently available deep-learning models, experimental results show that the proposed inception residual block can avoid the vanishing gradient problem and achieve a high accuracy rate of 92%.

AB - Ecologists have been studying different computational models in the classification of ecological species. In this paper, we intend to take advantages of variant deep-learning models, including LeNet, AlexNet, VGG models, residual neural network, and inception models, to classify ecological datasets, such as bee wing and butterfly. Since the datasets contain relatively small data samples and unbalanced samples in each class, we apply data augmentation and transfer learning techniques. Furthermore, newly designed inception residual and inception modules are developed to enhance feature extraction and increase classification rates. As comparing against currently available deep-learning models, experimental results show that the proposed inception residual block can avoid the vanishing gradient problem and achieve a high accuracy rate of 92%.

KW - Deep learning

KW - bee wings

KW - convolutional neural network

KW - ecology

KW - image augmentation

KW - image classification

KW - inception residual module

UR - http://www.scopus.com/inward/record.url?scp=85084395251&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85084395251&partnerID=8YFLogxK

U2 - https://doi.org/10.1142/S0218001420520102

DO - https://doi.org/10.1142/S0218001420520102

M3 - Article

SN - 0218-0014

VL - 34

JO - International Journal of Pattern Recognition and Artificial Intelligence

JF - International Journal of Pattern Recognition and Artificial Intelligence

IS - 13

M1 - 2052010

ER -

Classification of Ecological Data by Deep Learning

Abstract

ASJC Scopus subject areas

Keywords

Access to Document

Other files and links

Fingerprint

Cite this