Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions
In high-speed cigarette manufacturing industries, occasional minor cosmetic cigarette defects and a scarcity of samples significantly hinder the rapid and accurate detection of defects. To tackle this challenge, we propose an enhanced single-shot multibox detector (SSD) model that uses variational B...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIMS Press
2024-02-01
|
| Series: | Mathematical Biosciences and Engineering |
| Subjects: | |
| Online Access: | https://www.aimspress.com/article/doi/10.3934/mbe.2024145?viewType=HTML |
| _version_ | 1827340028451225600 |
|---|---|
| author | Shichao Wu Xianzhou Lv Yingbo Liu Ming Jiang Xingxu Li Dan Jiang Jing Yu Yunyu Gong Rong Jiang |
| author_facet | Shichao Wu Xianzhou Lv Yingbo Liu Ming Jiang Xingxu Li Dan Jiang Jing Yu Yunyu Gong Rong Jiang |
| author_sort | Shichao Wu |
| collection | DOAJ |
| description | In high-speed cigarette manufacturing industries, occasional minor cosmetic cigarette defects and a scarcity of samples significantly hinder the rapid and accurate detection of defects. To tackle this challenge, we propose an enhanced single-shot multibox detector (SSD) model that uses variational Bayesian inference for improved detection of tiny defects given sporadic occurrences and limited samples. The enhanced SSD model incorporates a bounded intersection over union (BIoU) loss function to reduce sensitivity to minor deviations and uses exponential linear unit (ELU) and leaky rectified linear unit (ReLU) activation functions to mitigate vanishing gradients and neuron death in deep neural networks. Empirical results show that the enhanced SSD300 and SSD512 models increase the model's detection accuracy mean average precision (mAP) by up to 1.2% for small defects. Ablation studies further reveal that the model's mAP increases by 1.5%, which reduces the computational requirements by 5.92 GFLOPs. The model also shows improved inference in scenarios with limited samples, thus highlighting its effectiveness and applicability in high-speed, precision-oriented cigarette manufacturing industries. |
| first_indexed | 2024-03-07T20:10:05Z |
| format | Article |
| id | doaj.art-9a02c09bf98b49cbb33148b5b83a8bf2 |
| institution | Directory Open Access Journal |
| issn | 1551-0018 |
| language | English |
| last_indexed | 2024-03-07T20:10:05Z |
| publishDate | 2024-02-01 |
| publisher | AIMS Press |
| record_format | Article |
| series | Mathematical Biosciences and Engineering |
| spelling | doaj.art-9a02c09bf98b49cbb33148b5b83a8bf22024-02-28T01:25:19ZengAIMS PressMathematical Biosciences and Engineering1551-00182024-02-012123281330310.3934/mbe.2024145Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditionsShichao Wu 0Xianzhou Lv1Yingbo Liu2Ming Jiang 3Xingxu Li4Dan Jiang 5Jing Yu6Yunyu Gong7Rong Jiang81. School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming 650221, China2. Hongyun Honghe Tobacco (Group) Co., Ltd. Huize Cigarette Factory, Qujing 654200, China1. School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming 650221, China2. Hongyun Honghe Tobacco (Group) Co., Ltd. Huize Cigarette Factory, Qujing 654200, China1. School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming 650221, China1. School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming 650221, China1. School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming 650221, China1. School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming 650221, China3. Yunnan Key Laboratory of Service Computing, Kunming 650221, ChinaIn high-speed cigarette manufacturing industries, occasional minor cosmetic cigarette defects and a scarcity of samples significantly hinder the rapid and accurate detection of defects. To tackle this challenge, we propose an enhanced single-shot multibox detector (SSD) model that uses variational Bayesian inference for improved detection of tiny defects given sporadic occurrences and limited samples. The enhanced SSD model incorporates a bounded intersection over union (BIoU) loss function to reduce sensitivity to minor deviations and uses exponential linear unit (ELU) and leaky rectified linear unit (ReLU) activation functions to mitigate vanishing gradients and neuron death in deep neural networks. Empirical results show that the enhanced SSD300 and SSD512 models increase the model's detection accuracy mean average precision (mAP) by up to 1.2% for small defects. Ablation studies further reveal that the model's mAP increases by 1.5%, which reduces the computational requirements by 5.92 GFLOPs. The model also shows improved inference in scenarios with limited samples, thus highlighting its effectiveness and applicability in high-speed, precision-oriented cigarette manufacturing industries.https://www.aimspress.com/article/doi/10.3934/mbe.2024145?viewType=HTMLssdcigarette appearance defectsvariational bayesian inferencetiny target detection |
| spellingShingle | Shichao Wu Xianzhou Lv Yingbo Liu Ming Jiang Xingxu Li Dan Jiang Jing Yu Yunyu Gong Rong Jiang Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions Mathematical Biosciences and Engineering ssd cigarette appearance defects variational bayesian inference tiny target detection |
| title | Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions |
| title_full | Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions |
| title_fullStr | Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions |
| title_full_unstemmed | Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions |
| title_short | Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions |
| title_sort | enhanced ssd framework for detecting defects in cigarette appearance using variational bayesian inference under limited sample conditions |
| topic | ssd cigarette appearance defects variational bayesian inference tiny target detection |
| url | https://www.aimspress.com/article/doi/10.3934/mbe.2024145?viewType=HTML |
| work_keys_str_mv | AT shichaowu enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT xianzhoulv enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT yingboliu enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT mingjiang enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT xingxuli enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT danjiang enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT jingyu enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT yunyugong enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions AT rongjiang enhancedssdframeworkfordetectingdefectsincigaretteappearanceusingvariationalbayesianinferenceunderlimitedsampleconditions |