Combining MobileNetV1 and Depthwise Separable convolution bottleneck with Expansion for classifying the freshness of fish eyes

Prasetyo, Eko and Purbaningtyas, Rani and Adityo, R Dimas and Suciati, Nanik and Fatichah, Chastine (2022) Combining MobileNetV1 and Depthwise Separable convolution bottleneck with Expansion for classifying the freshness of fish eyes. INFORMATION PROCESSING IN AGRICULTURE, 9 (4). pp. 485-496. ISSN 2214-3173

	Text IPA 2022 - Combining MobileNetV1 and Depthwise Separable convolution bottleneck with Expansion for classifying the freshness of f (4).pdf Download (1MB)
	Text Bukti Corr Author jurnal Information Processing in Agriculture.pdf Download (3MB)
	Text Berkas pendukung jurnal IPA.pdf Download (978kB)
	Text (cek plagiasi) Combining_MobileNetV1_and_Depthwise_Separable_conv (1).pdf Download (3MB)

Abstract

mage classification using Convolutional Neural Network (CNN) achieves optimal perfor�mance with a particular strategy. MobileNet reduces the parameter number for learnin features by switching from the standard convolution paradigm to the depthwise separable convolution (DSC) paradigm. However, there are not enough features to learn for identify�ing the freshness of fish eyes. Furthermore, minor variances in features should not require complicated CNN architecture. In this paper, our first contribution proposed DSC Bottle�neck with Expansion for learning features of the freshness of fish eyes with a Bottleneck Multiplier. The second contribution proposed Residual Transition to bridge current feature maps and skip connection feature maps to the next convolution block. The third contribu�tion proposed MobileNetV1 Bottleneck with Expansion (MB-BE) for classifying the freshness of fish eyes. The result obtained from the Freshness of the Fish Eyes dataset shows that MB�BE outperformed other models such as original MobileNet, VGG16, Densenet, Nasnet Mobile with 63.21% accuracy.

Item Type:	Article
Uncontrolled Keywords:	Depthwise separable convolution Bottleneck Classification Freshness Fish eye Residual transition
Subjects:	Technology
Divisions:	Faculty of Engineering
Depositing User:	Perpus Ubhara Surabaya
Date Deposited:	30 Apr 2024 06:07
Last Modified:	14 May 2024 04:37
URI:	http://eprints.ubhara.ac.id/id/eprint/2411

Actions (login required)

View Item