{"title":"Detection of explosives in dustbins using deep transfer learning based multiclass classifiers","authors":"Amoakoh Gyasi-Agyei","doi":"10.1007/s10489-023-05249-1","DOIUrl":null,"url":null,"abstract":"<h3 data-test=\"abstract-sub-heading\">Abstract</h3><p>The concealment of improvised explosive devices in dustbins aimed at destroying people and property is causing the mass removal of dustbins from public places and vehicular public transport in cities around the world. Such action of dustbin removal results in littering, stench, pests, contamination of water bodies, the spread of diseases, and increased greenhouse gases. The current solutions to the problem are blast-resistant dustbins which are bulky and expensive, and transparent dustbins which display the awful appearance of wastes in public places. This article proposes equipping dustbins with artificial intelligence-based classifiers to detect explosives concealed in wastes in public dustbins to minimise the risk to public safety. There was the need to construct a new database of explosive images to augment the existing TrashNet dataset. Then, through transfer learning using eight state-of-the-art convolutional neural networks as base models, the augmented dataset was used to search for optimum convolutional neural networks to detect explosives. One of the trained networks based on DenseNet-121 achieved the Top-1 accuracy of 80% with about 26 minutes learning time, which is 6.7% better than the Top-1 accuracy achieved by the base model on the benchmark ImageNet dataset. This finding demonstrates that the designed neural networks are promising cutting-edge techniques for detecting explosives concealed in dustbins to threaten public safety. To the best of our knowledge, this is the first time that convolutional neural networks have been proposed to identify explosives concealed in dustbins.</p><h3 data-test=\"abstract-sub-heading\">Graphical abstract</h3>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"21 1","pages":""},"PeriodicalIF":3.4000,"publicationDate":"2024-02-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Applied Intelligence","FirstCategoryId":"94","ListUrlMain":"https://doi.org/10.1007/s10489-023-05249-1","RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0
Abstract
The concealment of improvised explosive devices in dustbins aimed at destroying people and property is causing the mass removal of dustbins from public places and vehicular public transport in cities around the world. Such action of dustbin removal results in littering, stench, pests, contamination of water bodies, the spread of diseases, and increased greenhouse gases. The current solutions to the problem are blast-resistant dustbins which are bulky and expensive, and transparent dustbins which display the awful appearance of wastes in public places. This article proposes equipping dustbins with artificial intelligence-based classifiers to detect explosives concealed in wastes in public dustbins to minimise the risk to public safety. There was the need to construct a new database of explosive images to augment the existing TrashNet dataset. Then, through transfer learning using eight state-of-the-art convolutional neural networks as base models, the augmented dataset was used to search for optimum convolutional neural networks to detect explosives. One of the trained networks based on DenseNet-121 achieved the Top-1 accuracy of 80% with about 26 minutes learning time, which is 6.7% better than the Top-1 accuracy achieved by the base model on the benchmark ImageNet dataset. This finding demonstrates that the designed neural networks are promising cutting-edge techniques for detecting explosives concealed in dustbins to threaten public safety. To the best of our knowledge, this is the first time that convolutional neural networks have been proposed to identify explosives concealed in dustbins.
期刊介绍：
With a focus on research in artificial intelligence and neural networks, this journal addresses issues involving solutions of real-life manufacturing, defense, management, government and industrial problems which are too complex to be solved through conventional approaches and require the simulation of intelligent thought processes, heuristics, applications of knowledge, and distributed and parallel processing. The integration of these multiple approaches in solving complex problems is of particular importance.
The journal presents new and original research and technological developments, addressing real and complex issues applicable to difficult problems. It provides a medium for exchanging scientific research and technological achievements accomplished by the international community.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
