Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean Recognition

As global trade networks rapidly expand, coffee production and consumption have increased globally, profoundly influencing modern lifestyles. However, the coffee production process still demands substantial labor, especially in the selection and processing of coffee beans. The high implementation co...

Full description

Saved in:
Bibliographic Details
Main Authors: Liang-Ying Ke, Pin-Feng Lin, Chih-Hsien Hsia
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Engineering Proceedings
Subjects:
Online Access:https://www.mdpi.com/2673-4591/92/1/69
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849433861846466560
author Liang-Ying Ke
Pin-Feng Lin
Chih-Hsien Hsia
author_facet Liang-Ying Ke
Pin-Feng Lin
Chih-Hsien Hsia
author_sort Liang-Ying Ke
collection DOAJ
description As global trade networks rapidly expand, coffee production and consumption have increased globally, profoundly influencing modern lifestyles. However, the coffee production process still demands substantial labor, especially in the selection and processing of coffee beans. The high implementation costs have impeded its widespread adoption. Therefore, we developed a defect detection and roasting level recognition method using a lightweight vision transformer (ViT) based on the deep learning (DL) method to extract features from coffee bean images. The developed method effectively reduces the overall cost of the coffee production process, showing a recognition accuracy of 98.49% for the Coffee Cobra database and 99.68% for the Roasting Coffee Bean database. The number of the model parameters was only 0.13 M, making it appropriate to deploy to low-cost embedded platforms.
format Article
id doaj-art-d4b8000e84344f9fb8dca4fbbef84955
institution Kabale University
issn 2673-4591
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Engineering Proceedings
spelling doaj-art-d4b8000e84344f9fb8dca4fbbef849552025-08-20T03:26:52ZengMDPI AGEngineering Proceedings2673-45912025-05-019216910.3390/engproc2025092069Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean RecognitionLiang-Ying Ke0Pin-Feng Lin1Chih-Hsien Hsia2Department of Engineering Science, National Cheng Kung University, Tainan 701, TaiwanDa Ren Catholic High School, Taipei 114, TaiwanDepartment of Computer Science and Information Engineering, National Ilan University, Yilan 260, TaiwanAs global trade networks rapidly expand, coffee production and consumption have increased globally, profoundly influencing modern lifestyles. However, the coffee production process still demands substantial labor, especially in the selection and processing of coffee beans. The high implementation costs have impeded its widespread adoption. Therefore, we developed a defect detection and roasting level recognition method using a lightweight vision transformer (ViT) based on the deep learning (DL) method to extract features from coffee bean images. The developed method effectively reduces the overall cost of the coffee production process, showing a recognition accuracy of 98.49% for the Coffee Cobra database and 99.68% for the Roasting Coffee Bean database. The number of the model parameters was only 0.13 M, making it appropriate to deploy to low-cost embedded platforms.https://www.mdpi.com/2673-4591/92/1/69deep learningcoffee beanvision transformerlightweight
spellingShingle Liang-Ying Ke
Pin-Feng Lin
Chih-Hsien Hsia
Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean Recognition
Engineering Proceedings
deep learning
coffee bean
vision transformer
lightweight
title Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean Recognition
title_full Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean Recognition
title_fullStr Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean Recognition
title_full_unstemmed Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean Recognition
title_short Lightweight Depthwise Pooling Transformer for Enhanced Coffee Bean Recognition
title_sort lightweight depthwise pooling transformer for enhanced coffee bean recognition
topic deep learning
coffee bean
vision transformer
lightweight
url https://www.mdpi.com/2673-4591/92/1/69
work_keys_str_mv AT liangyingke lightweightdepthwisepoolingtransformerforenhancedcoffeebeanrecognition
AT pinfenglin lightweightdepthwisepoolingtransformerforenhancedcoffeebeanrecognition
AT chihhsienhsia lightweightdepthwisepoolingtransformerforenhancedcoffeebeanrecognition