Research on Food Image Classification Algorithm based on Improved MobileNetV3-Large

Food image recognition plays a crucial role in food safety monitoring, nutritional analysis, and dietary recommendation systems. However, the diversity, complexity, and external factors such as lighting conditions pose numerous difficulties and challenges to the recognition task. In order to address...

Full description

Saved in:
Bibliographic Details
Main Authors: HE Wei-chan, YANG Zhi-jing, QIN Jing-hui
Format: Article
Language:English
Published: Academy of National Food and Strategic Reserves Administration 2025-03-01
Series:Liang you shipin ke-ji
Subjects:
Online Access:http://lyspkj.ijournal.cn/lyspkj/article/abstract/20250214?st=article_issue
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849337900387270656
author HE Wei-chan
YANG Zhi-jing
QIN Jing-hui
author_facet HE Wei-chan
YANG Zhi-jing
QIN Jing-hui
author_sort HE Wei-chan
collection DOAJ
description Food image recognition plays a crucial role in food safety monitoring, nutritional analysis, and dietary recommendation systems. However, the diversity, complexity, and external factors such as lighting conditions pose numerous difficulties and challenges to the recognition task. In order to address these issues, this paper proposed a food image classification algorithm based on improved MobileNetV3-Large. Firstly, building upon the pre-trained MobileNetV3-Large model, the PReLu activation function and NAM attention mechanism were introduced to enhance the model's focus on key features by capturing non-local dependencies in images. Subsequently, a multi-task loss function was incorporated to further improve the classification performance by simultaneously optimizing multiple related tasks. Finally, the TrivialAugment data augmentation technique was employed to enhance the model's generalization ability by expanding the scale and diversity of the training dataset. Experimental results demonstrated that through these improvements, the model's accuracy on the Food-101 dataset increased from 66.9% to 84.2%, demonstrating the effectiveness of the proposed approach.
format Article
id doaj-art-54dbd9b4c7e44ad9ab5e948ead6fe2c7
institution Kabale University
issn 1007-7561
language English
publishDate 2025-03-01
publisher Academy of National Food and Strategic Reserves Administration
record_format Article
series Liang you shipin ke-ji
spelling doaj-art-54dbd9b4c7e44ad9ab5e948ead6fe2c72025-08-20T03:44:33ZengAcademy of National Food and Strategic Reserves AdministrationLiang you shipin ke-ji1007-75612025-03-01332909610.16210/j.cnki.1007-7561.2025.02.013Research on Food Image Classification Algorithm based on Improved MobileNetV3-LargeHE Wei-chan0YANG Zhi-jing1QIN Jing-hui2School of Information Engineering, Guangdong University of Technology, Guangzhou, Guangdong 510000, ChinaSchool of Information Engineering, Guangdong University of Technology, Guangzhou, Guangdong 510000, ChinaSchool of Information Engineering, Guangdong University of Technology, Guangzhou, Guangdong 510000, ChinaFood image recognition plays a crucial role in food safety monitoring, nutritional analysis, and dietary recommendation systems. However, the diversity, complexity, and external factors such as lighting conditions pose numerous difficulties and challenges to the recognition task. In order to address these issues, this paper proposed a food image classification algorithm based on improved MobileNetV3-Large. Firstly, building upon the pre-trained MobileNetV3-Large model, the PReLu activation function and NAM attention mechanism were introduced to enhance the model's focus on key features by capturing non-local dependencies in images. Subsequently, a multi-task loss function was incorporated to further improve the classification performance by simultaneously optimizing multiple related tasks. Finally, the TrivialAugment data augmentation technique was employed to enhance the model's generalization ability by expanding the scale and diversity of the training dataset. Experimental results demonstrated that through these improvements, the model's accuracy on the Food-101 dataset increased from 66.9% to 84.2%, demonstrating the effectiveness of the proposed approach.http://lyspkj.ijournal.cn/lyspkj/article/abstract/20250214?st=article_issuemobilenetv3-largenam attention mechanismprelu activation functiontrivialaugment data augmentation
spellingShingle HE Wei-chan
YANG Zhi-jing
QIN Jing-hui
Research on Food Image Classification Algorithm based on Improved MobileNetV3-Large
Liang you shipin ke-ji
mobilenetv3-large
nam attention mechanism
prelu activation function
trivialaugment data augmentation
title Research on Food Image Classification Algorithm based on Improved MobileNetV3-Large
title_full Research on Food Image Classification Algorithm based on Improved MobileNetV3-Large
title_fullStr Research on Food Image Classification Algorithm based on Improved MobileNetV3-Large
title_full_unstemmed Research on Food Image Classification Algorithm based on Improved MobileNetV3-Large
title_short Research on Food Image Classification Algorithm based on Improved MobileNetV3-Large
title_sort research on food image classification algorithm based on improved mobilenetv3 large
topic mobilenetv3-large
nam attention mechanism
prelu activation function
trivialaugment data augmentation
url http://lyspkj.ijournal.cn/lyspkj/article/abstract/20250214?st=article_issue
work_keys_str_mv AT heweichan researchonfoodimageclassificationalgorithmbasedonimprovedmobilenetv3large
AT yangzhijing researchonfoodimageclassificationalgorithmbasedonimprovedmobilenetv3large
AT qinjinghui researchonfoodimageclassificationalgorithmbasedonimprovedmobilenetv3large