{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T18:15:54Z","timestamp":1784139354273,"version":"3.55.0"},"reference-count":38,"publisher":"Public Library of Science (PLoS)","issue":"11","license":[{"start":{"date-parts":[[2025,11,21]],"date-time":"2025-11-21T00:00:00Z","timestamp":1763683200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100006407","name":"Natural Science Foundation of Henan Province","doi-asserted-by":"publisher","award":["242300421718 (Youth Foundation)"],"award-info":[{"award-number":["242300421718 (Youth Foundation)"]}],"id":[{"id":"10.13039\/501100006407","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100006407","name":"Natural Science Foundation of Henan Province","doi-asserted-by":"publisher","award":["252300420399 (Superficial Foundation)"],"award-info":[{"award-number":["252300420399 (Superficial Foundation)"]}],"id":[{"id":"10.13039\/501100006407","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Quzhou City Science and Technology Plan","award":["2023K044"],"award-info":[{"award-number":["2023K044"]}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>Vehicle classification is a core task in intelligent transportation systems, where high demands are placed on both computational efficiency and generalization ability in practical applications. Existing deep learning models often struggle to meet these requirements due to their high computational complexity and limited generalization. To address this challenge, this study proposes a lightweight and efficient deep neural network called DSICBAMNet, which achieves high classification accuracy while significantly improving computational efficiency. The design of DSICBAMNet is centered on two key components: Depthwise Separable Convolutions (DSC) and an Improved Convolutional Block Attention Module (ICBAM). The DSC module reduces the number of parameters and computational complexity by decomposing convolution operations, making it well-suited for resource-constrained deployment scenarios. Meanwhile, ICBAM addresses the shortcomings of traditional CBAM in terms of overfitting resistance and feature weighting strategies. By introducing Dropout regularization into the channel attention module, ICBAM enhances the model\u2019s resistance to overfitting. Additionally, it optimizes the interaction mechanisms and weight distribution between the channel and spatial attention modules, enabling more accurate multi-class feature representation. The network achieves efficient multi-scale feature extraction by stacking multiple improved DSICBAM blocks while maintaining an overall lightweight structure. In experimental evaluations, DSICBAMNet was compared with five classic models, including AlexNet and MobileNetV2. Experimental results demonstrate that DSICBAMNet achieves outstanding performance on both the MIO-TCD dataset, with 286 test samples and an average classification accuracy of 97.36%, and the Stanford Cars dataset, with 1,060 test samples and an accuracy of 96.51%. Moreover, the combination of Grad-CAM visualizations and confusion matrix analysis validates the model\u2019s ability to focus on key regions and maintain consistency in classification outcomes. These results underscore the model\u2019s potential applicability and practical value in intelligent transportation scenarios.<\/jats:p>","DOI":"10.1371\/journal.pone.0335967","type":"journal-article","created":{"date-parts":[[2025,11,21]],"date-time":"2025-11-21T18:35:36Z","timestamp":1763750136000},"page":"e0335967","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":3,"title":["Lightweight deep neural networks: Optimization of vehicle classification using ICBAM based on depthwise separable convolutions"],"prefix":"10.1371","volume":"20","author":[{"given":"Qifeng","family":"Niu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jinhui","family":"Han","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-8836-6938","authenticated-orcid":true,"given":"Zhen","family":"Sui","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Feng","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2025,11,21]]},"reference":[{"key":"pone.0335967.ref001","doi-asserted-by":"crossref","first-page":"173","DOI":"10.1109\/OJITS.2021.3096756","article-title":"Vehicle classification in intelligent transport systems: an overview, methods and software perspective","volume":"2","author":"A Gholamhosseinian","year":"2021","journal-title":"IEEE Open J Intell Transp Syst."},{"issue":"1","key":"pone.0335967.ref002","doi-asserted-by":"crossref","first-page":"77","DOI":"10.1109\/MVT.2009.935537","article-title":"Intelligent transportation systems","volume":"5","author":"G Dimitrakopoulos","year":"2010","journal-title":"IEEE Veh Technol Mag."},{"key":"pone.0335967.ref003","doi-asserted-by":"crossref","first-page":"14643","DOI":"10.1109\/ACCESS.2022.3145972","article-title":"On the integration of enabling wireless technologies and sensor fusion for next-generation connected and autonomous vehicles","volume":"10","author":"FA Butt","year":"2022","journal-title":"IEEE Access."},{"issue":"2","key":"pone.0335967.ref004","doi-asserted-by":"crossref","first-page":"3140","DOI":"10.1109\/TIV.2023.3339798","article-title":"Intelligent cockpit for intelligent connected vehicles: definition, taxonomy, technology and evaluation","volume":"9","author":"W Li","year":"2024","journal-title":"IEEE Trans Intell Veh."},{"issue":"3","key":"pone.0335967.ref005","doi-asserted-by":"crossref","first-page":"508","DOI":"10.1109\/TCSVT.2014.2358031","article-title":"Efficient feature selection and classification for vehicle detection","volume":"25","author":"X Wen","year":"2015","journal-title":"IEEE Trans Circuits Syst Video Technol."},{"issue":"4","key":"pone.0335967.ref006","doi-asserted-by":"crossref","first-page":"1077","DOI":"10.1007\/s12065-022-00723-0","article-title":"Automatic vehicle detection system in day and night mode: challenges, applications and panoramic review","volume":"16","author":"N Arora","year":"2022","journal-title":"Evol Intel."},{"issue":"1","key":"pone.0335967.ref007","doi-asserted-by":"crossref","first-page":"109","DOI":"10.1007\/s11370-022-00452-4","article-title":"Unmanned aerial vehicles (UAVs): practical aspects, applications, open challenges, security issues, and future trends","volume":"16","author":"SAH Mohsan","year":"2023","journal-title":"Intell Serv Robot."},{"issue":"9","key":"pone.0335967.ref008","doi-asserted-by":"crossref","first-page":"2352","DOI":"10.1162\/neco_a_00990","article-title":"Deep convolutional neural networks for image classification: a comprehensive review","volume":"29","author":"W Rawat","year":"2017","journal-title":"Neural Comput."},{"issue":"04","key":"pone.0335967.ref009","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1117\/1.JRS.11.042609","article-title":"Comprehensive survey of deep learning in remote sensing: theories, tools, and challenges for the community","volume":"11","author":"JE Ball","year":"2017","journal-title":"J Appl Remote Sens."},{"issue":"11","key":"pone.0335967.ref010","doi-asserted-by":"crossref","first-page":"6068","DOI":"10.1109\/TNNLS.2021.3082304","article-title":"Custom hardware architectures for deep learning on portable devices: a review","volume":"33","author":"KS Zaman","year":"2022","journal-title":"IEEE Trans Neural Netw Learn Syst."},{"issue":"2","key":"pone.0335967.ref011","doi-asserted-by":"crossref","first-page":"1241","DOI":"10.1109\/TIV.2024.3425647","article-title":"A review of personalization in driving behavior: dataset, modeling, and validation","volume":"10","author":"X Liao","year":"2025","journal-title":"IEEE Trans Intell Veh."},{"key":"pone.0335967.ref012","doi-asserted-by":"crossref","first-page":"123826","DOI":"10.1016\/j.energy.2022.123826","article-title":"Generalization ability of hybrid electric vehicle energy management strategy based on reinforcement learning method","volume":"250","author":"C Qi","year":"2022","journal-title":"Energy."},{"key":"pone.0335967.ref013","doi-asserted-by":"crossref","first-page":"19626","DOI":"10.1109\/ACCESS.2018.2813079","article-title":"Classifiers combination techniques: a comprehensive review","volume":"6","author":"M Mohandes","year":"2018","journal-title":"IEEE Access."},{"issue":"1","key":"pone.0335967.ref014","doi-asserted-by":"crossref","first-page":"1","DOI":"10.26599\/JICV.2023.9210019","article-title":"A review of vehicle detection methods based on computer vision","volume":"7","author":"C Ma","year":"2024","journal-title":"J Int Con Veh."},{"issue":"3","key":"pone.0335967.ref015","doi-asserted-by":"crossref","first-page":"1288","DOI":"10.1109\/TITS.2019.2906821","article-title":"Accurate classification for automatic vehicle-type recognition based on ensemble classifiers","volume":"21","author":"N Shvai","year":"2020","journal-title":"IEEE Trans Intell Transport Syst."},{"issue":"17","key":"pone.0335967.ref016","doi-asserted-by":"crossref","first-page":"2746","DOI":"10.3390\/diagnostics13172746","article-title":"A modified LeNet CNN for breast cancer diagnosis in ultrasound images","volume":"13","author":"S Balasubramaniam","year":"2023","journal-title":"Diagnostics (Basel)."},{"issue":"6","key":"pone.0335967.ref017","doi-asserted-by":"crossref","first-page":"1936","DOI":"10.1007\/s10618-020-00710-y","article-title":"InceptionTime: finding AlexNet for time series classification","volume":"34","author":"H Ismail Fawaz","year":"2020","journal-title":"Data Min Knowl Disc."},{"key":"pone.0335967.ref018","doi-asserted-by":"crossref","first-page":"95","DOI":"10.3389\/fnins.2019.00095","article-title":"Going deeper in spiking neural networks: VGG and residual architectures","volume":"13","author":"A Sengupta","year":"2019","journal-title":"Frontiers in Neuroscience."},{"issue":"10","key":"pone.0335967.ref019","doi-asserted-by":"crossref","first-page":"6111","DOI":"10.1007\/s00521-019-04097-w","article-title":"A transfer convolutional neural network for fault diagnosis based on ResNet-50","volume":"32","author":"L Wen","year":"2019","journal-title":"Neural Comput & Applic."},{"issue":"3","key":"pone.0335967.ref020","doi-asserted-by":"crossref","first-page":"1906","DOI":"10.3390\/su15031906","article-title":"Fruit image classification model based on MobileNetV2 with deep transfer learning technique","volume":"15","author":"Y Gulzar","year":"2023","journal-title":"Sustainability."},{"key":"pone.0335967.ref021","article-title":"S\u0131n\u0131fland\u0131rma Algoritmalar\u0131na Dayal\u0131 VGG-11 ile Y\u00fczde Duygu Tan\u0131ma","author":"K Donuk","year":"2021","journal-title":"Computer Science."},{"issue":"17","key":"pone.0335967.ref022","doi-asserted-by":"crossref","first-page":"2752","DOI":"10.3390\/diagnostics13172752","article-title":"DSCNet: deep skip connections-based dense network for ALL diagnosis using peripheral blood smear images","volume":"13","author":"M Kaur","year":"2023","journal-title":"Diagnostics (Basel)."},{"key":"pone.0335967.ref023","doi-asserted-by":"crossref","first-page":"463","DOI":"10.1016\/j.procs.2020.06.115","article-title":"Facial expression recognition via a CBAM embedded network","volume":"174","author":"W Cao","year":"2020","journal-title":"Procedia Computer Science."},{"issue":"5","key":"pone.0335967.ref024","first-page":"2280","article-title":"ACE-CNN: approximate carry disregard multipliers for energy-efficient CNN-based image classification","volume":"71","author":"S Shakibhamedan","year":"2024","journal-title":"IEEE Trans Circuits Syst I."},{"key":"pone.0335967.ref025","unstructured":"Tatsunami Y, Taki M. Sequencer: deep LSTM for image classification. In: Advances in Neural Information Processing Systems. 2022. p. 38204\u201317."},{"key":"pone.0335967.ref026","unstructured":"Goerttler S, Wang Y, Eldele E, Wu M, He F. MSA-CNN: a lightweight multi-scale CNN with attention for sleep stage classification. arXiv preprint 2025. arXiv:2501.02949"},{"issue":"5","key":"pone.0335967.ref027","doi-asserted-by":"crossref","first-page":"2689","DOI":"10.1007\/s11042-024-20300-0","article-title":"Enhanced satellite imagery analysis for post-disaster building damage assessment using integrated ResNet-U-Net model","volume":"84","author":"D Bhardwaj","year":"2024","journal-title":"Multimed Tools Appl."},{"issue":"1","key":"pone.0335967.ref028","doi-asserted-by":"crossref","first-page":"599","DOI":"10.1007\/s41870-024-02081-9","article-title":"Designing algorithm for context based analysis using deep learning (CNN+RNN) with image dataset","volume":"17","author":"AD Gawali","year":"2024","journal-title":"Int J Inf Tecnol."},{"key":"pone.0335967.ref029","unstructured":"Li Y, Yuan G, Wen Y, Hu J, Evangelidis G, Tulyakov S. Efficientformer: vision transformers at mobilenet speed. In: Advances in Neural Information Processing Systems, 2022. 12934\u201349."},{"key":"pone.0335967.ref030","doi-asserted-by":"crossref","first-page":"6817","DOI":"10.1109\/JSTARS.2022.3198517","article-title":"A CBAM based multiscale transformer fusion approach for remote sensing image change detection","volume":"15","author":"W Wang","year":"2022","journal-title":"IEEE J Sel Top Appl Earth Observations Remote Sensing."},{"key":"pone.0335967.ref031","doi-asserted-by":"crossref","first-page":"107683","DOI":"10.1016\/j.eiar.2024.107683","article-title":"Implementation of the EU carbon border adjustment mechanism and China\u2019s policy and legal responses","volume":"110","author":"J Chang","year":"2025","journal-title":"Environmental Impact Assessment Review."},{"issue":"10","key":"pone.0335967.ref032","doi-asserted-by":"crossref","first-page":"5129","DOI":"10.1109\/TIP.2018.2848705","article-title":"MIO-TCD: a new benchmark dataset for vehicle classification and localization","volume":"27","author":"Z Luo","year":"2018","journal-title":"IEEE Transactions on Image Processing."},{"issue":"2","key":"pone.0335967.ref033","doi-asserted-by":"crossref","first-page":"123","DOI":"10.1109\/81.372854","article-title":"Chua\u2019s circuit can be generated by CNN cells","volume":"42","author":"P Arena","year":"1995","journal-title":"IEEE Trans Circuits Syst I."},{"issue":"6","key":"pone.0335967.ref034","doi-asserted-by":"crossref","first-page":"951","DOI":"10.3390\/electronics11060951","article-title":"AlexNet convolutional neural network for disease detection and classification of tomato leaf","volume":"11","author":"H-C Chen","year":"2022","journal-title":"Electronics."},{"issue":"17","key":"pone.0335967.ref035","doi-asserted-by":"crossref","first-page":"6341","DOI":"10.3390\/en15176341","article-title":"Photovoltaic array reconfiguration under partial shading conditions based on short-circuit current estimated by convolutional neural network","volume":"15","author":"T Nguyen-Duc","year":"2022","journal-title":"Energies."},{"key":"pone.0335967.ref036","doi-asserted-by":"crossref","first-page":"83","DOI":"10.1186\/s13007-020-00624-2","article-title":"Early recognition of tomato gray leaf spot disease based on MobileNetv2-YOLOv3 model","volume":"16","author":"J Liu","year":"2020","journal-title":"Plant Methods."},{"key":"pone.0335967.ref037","doi-asserted-by":"crossref","first-page":"123061","DOI":"10.1016\/j.eswa.2023.123061","article-title":"A dense-sparse complementary network for human action recognition based on RGB and skeleton modalities","volume":"244","author":"Q Cheng","year":"2024","journal-title":"Expert Systems with Applications."},{"key":"pone.0335967.ref038","doi-asserted-by":"crossref","first-page":"105550","DOI":"10.1016\/j.compbiomed.2022.105550","article-title":"Explainable detection of myocardial infarction using deep learning models with Grad-CAM technique on ECG signals","volume":"146","author":"V Jahmunah","year":"2022","journal-title":"Comput Biol Med."}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0335967","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,21]],"date-time":"2025-11-21T18:35:45Z","timestamp":1763750145000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0335967"}},"subtitle":[],"editor":[{"given":"Ping","family":"Xiang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2025,11,21]]},"references-count":38,"journal-issue":{"issue":"11","published-online":{"date-parts":[[2025,11,21]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0335967","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,21]]}}}