Multi-level residual network VGGNet for fish species classification

The development of an image-based fish classification system using Convolutional Neural Network (CNN) has the advantages of no longer directly conducting features extraction and several features analysis. These steps has been involved by cascading convolution from initial to final block, where the i...

Full description

Saved in:
Bibliographic Details
Main Authors: Eko Prasetyo, Nanik Suciati, Chastine Fatichah
Format: Article
Language:English
Published: Springer 2022-09-01
Series:Journal of King Saud University: Computer and Information Sciences
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1319157821001300
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849315870034100224
author Eko Prasetyo
Nanik Suciati
Chastine Fatichah
author_facet Eko Prasetyo
Nanik Suciati
Chastine Fatichah
author_sort Eko Prasetyo
collection DOAJ
description The development of an image-based fish classification system using Convolutional Neural Network (CNN) has the advantages of no longer directly conducting features extraction and several features analysis. These steps has been involved by cascading convolution from initial to final block, where the initial, middle, and final blocks produce low-, middle-, and high-level features, respectively. Due to cascading convolution, CNN produces only high-level features. However, fish classification requires not only high-level features but also low-level features such as points, lines, and textures for representing edge spines, gill covers, fins, and skin textures in order to achieve higher performance; furthermore, CNN generally has not yet incorporated low-level features in the last block. In this paper, we proposed Multi-Level Residual (MLR) as a new residual network strategy by combining low-level features of the initial block with high-level features of the last block by applying Depthwise Separable Convolution. We also proposed MLR-VGGNet as a new CNN architecture inherited from VGGNet and strengthened it using Asymmetric Convolution, MLR, Batch Normalization, and Residual features. Our experimental results show that MLR-VGGNet achieved an accuracy of 99.69%, outperformed original VGGNet relative up to 10.33% and other CNN models relative up to 5.24% on Fish-gres and Fish4-Knowledge dataset.
format Article
id doaj-art-e4924bc3b4c94f67a1ac0e9a8c9fbedc
institution Kabale University
issn 1319-1578
language English
publishDate 2022-09-01
publisher Springer
record_format Article
series Journal of King Saud University: Computer and Information Sciences
spelling doaj-art-e4924bc3b4c94f67a1ac0e9a8c9fbedc2025-08-20T03:52:02ZengSpringerJournal of King Saud University: Computer and Information Sciences1319-15782022-09-013485286529510.1016/j.jksuci.2021.05.015Multi-level residual network VGGNet for fish species classificationEko Prasetyo0Nanik Suciati1Chastine Fatichah2Department of Informatics, Faculty of Intelligent Electrical and Informatics Technology, Institut Teknologi Sepuluh Nopember, Jl. Raya ITS, Surabaya 60111, Indonesia; Department of Informatics, Faculty of Engineering, Universitas Bhayangkara Surabaya, Jl. Ahmad Yani 114, Surabaya 60231, IndonesiaDepartment of Informatics, Faculty of Intelligent Electrical and Informatics Technology, Institut Teknologi Sepuluh Nopember, Jl. Raya ITS, Surabaya 60111, Indonesia; Corresponding author.Department of Informatics, Faculty of Intelligent Electrical and Informatics Technology, Institut Teknologi Sepuluh Nopember, Jl. Raya ITS, Surabaya 60111, IndonesiaThe development of an image-based fish classification system using Convolutional Neural Network (CNN) has the advantages of no longer directly conducting features extraction and several features analysis. These steps has been involved by cascading convolution from initial to final block, where the initial, middle, and final blocks produce low-, middle-, and high-level features, respectively. Due to cascading convolution, CNN produces only high-level features. However, fish classification requires not only high-level features but also low-level features such as points, lines, and textures for representing edge spines, gill covers, fins, and skin textures in order to achieve higher performance; furthermore, CNN generally has not yet incorporated low-level features in the last block. In this paper, we proposed Multi-Level Residual (MLR) as a new residual network strategy by combining low-level features of the initial block with high-level features of the last block by applying Depthwise Separable Convolution. We also proposed MLR-VGGNet as a new CNN architecture inherited from VGGNet and strengthened it using Asymmetric Convolution, MLR, Batch Normalization, and Residual features. Our experimental results show that MLR-VGGNet achieved an accuracy of 99.69%, outperformed original VGGNet relative up to 10.33% and other CNN models relative up to 5.24% on Fish-gres and Fish4-Knowledge dataset.http://www.sciencedirect.com/science/article/pii/S1319157821001300Multi-level residualLow level featureConvolutional neural networkAsymmetric convolutionFish species classificationVGGNet
spellingShingle Eko Prasetyo
Nanik Suciati
Chastine Fatichah
Multi-level residual network VGGNet for fish species classification
Journal of King Saud University: Computer and Information Sciences
Multi-level residual
Low level feature
Convolutional neural network
Asymmetric convolution
Fish species classification
VGGNet
title Multi-level residual network VGGNet for fish species classification
title_full Multi-level residual network VGGNet for fish species classification
title_fullStr Multi-level residual network VGGNet for fish species classification
title_full_unstemmed Multi-level residual network VGGNet for fish species classification
title_short Multi-level residual network VGGNet for fish species classification
title_sort multi level residual network vggnet for fish species classification
topic Multi-level residual
Low level feature
Convolutional neural network
Asymmetric convolution
Fish species classification
VGGNet
url http://www.sciencedirect.com/science/article/pii/S1319157821001300
work_keys_str_mv AT ekoprasetyo multilevelresidualnetworkvggnetforfishspeciesclassification
AT naniksuciati multilevelresidualnetworkvggnetforfishspeciesclassification
AT chastinefatichah multilevelresidualnetworkvggnetforfishspeciesclassification