{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,11]],"date-time":"2026-08-11T11:37:03Z","timestamp":1786448223204,"version":"build-2736575974"},"reference-count":56,"publisher":"Springer Science and Business Media LLC","issue":"6","license":[{"start":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T00:00:00Z","timestamp":1782777600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,8,11]],"date-time":"2026-08-11T00:00:00Z","timestamp":1786406400000},"content-version":"vor","delay-in-days":42,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J. King Saud Univ. Comput. Inf. Sci."],"published-print":{"date-parts":[[2026,8]]},"abstract":"Abstract<\/jats:title>\n \n Rubber tree (\n Hevea brasiliensis<\/jats:italic>\n ) powdery mildew (PM) is a devastating foliar disease causing annual yield losses exceeding 30% in severely affected plantations. Automated multi-grade detection in complex field conditions is hampered by three technical gaps: (i) CNN-based detectors lack global context for disambiguating fine-grained early lesions; (ii) Transformer detectors incur quadratic complexity\n \n \n $$O(n^2)$$<\/jats:tex-math>\n \n \n O<\/mml:mi>\n (<\/mml:mo>\n \n n<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n )<\/mml:mo>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:alternatives>\n <\/jats:inline-formula>\n impractical for edge deployment; and (iii) conventional multi-scale fusion inadequately bridges the semantic gap across disease scales. We propose\n LightPM-DETR<\/jats:bold>\n , a lightweight detection transformer addressing these gaps through a co-designed architecture\u2013compression pipeline. The encoder incorporates four modules: (1)\n Selective Feature Scanning (SFS)<\/jats:bold>\n , a newly designed tri-stream Mamba adaptation for\n O<\/jats:italic>\n (\n n<\/jats:italic>\n )-complexity long-range modelling; (2)\n Poly-Kernel Inception Net (PKI Net)<\/jats:bold>\n , adapted for multi-scale receptive field aggregation; (3)\n Wavelet Efficient Attention (WEA)<\/jats:bold>\n , a new Haar-basis frequency-domain key-value compression mechanism for lesion-sensitive attention; and (4)\n Global Query Aggregator (GQA)<\/jats:bold>\n , a new hierarchical cross-attention module for query enrichment. Structured channel pruning followed by IoU-weighted knowledge distillation further compresses the model by 85.3% in parameters relative to RT-DETR-L (from 32.7\u00a0M to 4.8\u00a0M). On the self-built\n PM-Dataset-Plus<\/jats:bold>\n (6-grade, 8,412 images, 47,836 instances) and\n PD40<\/jats:bold>\n (40-class cross-species) benchmarks, LightPM-DETR achieves mAP\n \n \n $$_{50}$$<\/jats:tex-math>\n \n \n \n 50<\/mml:mn>\n \n <\/mml:mmultiscripts>\n <\/mml:math>\n <\/jats:alternatives>\n <\/jats:inline-formula>\n of\n 91.6%<\/jats:bold>\n and\n 84.3%<\/jats:bold>\n respectively (mean over three runs), surpassing all baselines. The compressed variant operates at 4.8\u00a0M parameters and 6.2\u00a0G FLOPs, achieving 62.4\u00a0FPS on a desktop GPU and 23.8\u00a0FPS on an NVIDIA Jetson Orin Nano. Code and data will be released at\n https:\/\/github.com\/wfcyliyuheng-dev\/PM-Dataset-PLus<\/jats:ext-link>\n upon acceptance.\n <\/jats:p>","DOI":"10.1007\/s44443-026-00958-2","type":"journal-article","created":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T19:37:50Z","timestamp":1782848270000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["LightPM-DETR: a lightweight transformer for grading detection of rubber tree powdery mildew"],"prefix":"10.1007","volume":"38","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-8690-4680","authenticated-orcid":false,"given":"Licheng","family":"Zhang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-1849-2728","authenticated-orcid":false,"given":"Yuheng","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,6,30]]},"reference":[{"key":"958_CR1","doi-asserted-by":"publisher","first-page":"96","DOI":"10.1016\/j.biosystemseng.2019.02.002","volume":"180","author":"JGA Barbedo","year":"2019","unstructured":"Barbedo JGA (2019) Plant disease identification from individual lesions and spots using deep learning. Biosys Eng 180:96\u2013107","journal-title":"Biosys Eng"},{"key":"958_CR2","unstructured":"Bochkovskiy A, Wang C-Y, Liao H-YM (2020) YOLOv4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934"},{"key":"958_CR3","doi-asserted-by":"crossref","unstructured":"Carion N, Massa F, Synnaeve G, Usunier N, Kirillov A, Zagoruyko S (2020) End-to-end object detection with transformers. In: European conference on computer vision, pp 213\u2013229. Springer","DOI":"10.1007\/978-3-030-58452-8_13"},{"key":"958_CR4","unstructured":"Chen C, Liu Y, Dong Y, Han J et al (2023) PKINet: Rethinking backbone design for aerial image object detection. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 17574\u201317583"},{"key":"958_CR5","unstructured":"Chen S, Zhang W, Liu J, Hu J (2024a) Lightweight plant disease detection via multi-scale feature aggregation and transformer encoder. Comput Electron Agric 216:108524"},{"key":"958_CR6","doi-asserted-by":"crossref","unstructured":"Chen K, Chen B, Liu C, Li W, Zou Z, Shi Z (2024b) RSMamba: Remote sensing image classification with state space model. IEEE Geosci Remote Sens Lett","DOI":"10.1109\/LGRS.2024.3407111"},{"issue":"3","key":"958_CR7","doi-asserted-by":"crossref","first-page":"447","DOI":"10.3390\/agriculture14030447","volume":"14","author":"VQ Dinh","year":"2024","unstructured":"Dinh VQ, Nguyen TTH et al (2024) Smart monitoring system for rubber tree disease using deep learning. Agriculture 14(3):447","journal-title":"Agriculture"},{"key":"958_CR8","doi-asserted-by":"publisher","first-page":"311","DOI":"10.1016\/j.compag.2018.01.009","volume":"145","author":"KP Ferentinos","year":"2018","unstructured":"Ferentinos KP (2018) Deep learning models for plant disease detection and diagnosis. Comput Electron Agric 145:311\u2013318","journal-title":"Comput Electron Agric"},{"key":"958_CR9","unstructured":"Frankle J, Carlin M (2019) The lottery ticket hypothesis: Finding sparse, trainable neural networks. In: International conference on learning representations"},{"key":"958_CR10","unstructured":"Gu A, Dao T (2023) Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752"},{"key":"958_CR11","first-page":"35971","volume":"35","author":"A Gu","year":"2022","unstructured":"Gu A, Goel K, Gupta A, Re C (2022) On the parameterization and initialization of diagonal state space models 35:35971\u201335983","journal-title":"On the parameterization and initialization of diagonal state space models"},{"key":"958_CR12","unstructured":"Gu A, Goel K, Re C (2021) Efficiently modeling long sequences with structured state spaces. arXiv preprint arXiv:2111.00396"},{"key":"958_CR13","unstructured":"Han S, Pool J, Tran J, Dally W (2015) Learning both weights and connections for efficient neural networks. In: Advances in neural information processing systems, vol 28"},{"key":"958_CR14","unstructured":"Hinton G, Vinyals O, Dean J (2015) Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531"},{"key":"958_CR15","unstructured":"Howard AG, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, Andreetto M, Adam H (2017) MobileNets: Efficient convolutional neural networks for mobile vision applications. In: arXiv Preprint arXiv:1704.04861"},{"key":"958_CR16","unstructured":"Hughes DP, Salathe M (2015) PlantVillage dataset. arXiv preprint arXiv:1511.08060"},{"key":"958_CR17","doi-asserted-by":"crossref","unstructured":"Hu J, Shen L, Sun G (2018) Squeeze-and-excitation networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 7132\u20137141","DOI":"10.1109\/CVPR.2018.00745"},{"key":"958_CR18","volume":"7","author":"T Jayasinghe","year":"2024","unstructured":"Jayasinghe T et al (2024) A deep learning framework for rubber tree disease detection using uav imagery. Smart Agricultural Technology 7:100378","journal-title":"Smart Agricultural Technology"},{"key":"958_CR19","unstructured":"Jocher G, Chaurasia A, Qiu J (2023) Ultralytics YOLOv8. GitHub repository"},{"issue":"6","key":"958_CR20","doi-asserted-by":"publisher","first-page":"671","DOI":"10.3390\/horticulturae11060671","volume":"11","author":"F Khan","year":"2025","unstructured":"Khan F et al (2025) Lightweight multi-scale detection of field crop pests and diseases for agricultural robots. Horticulturae 11(6):671. https:\/\/doi.org\/10.3390\/horticulturae11060671","journal-title":"Horticulturae"},{"key":"958_CR21","unstructured":"Khanam R, Hussain M (2024) YOLOv11: An overview of the key architectural enhancements. arXiv preprint arXiv:2410.17725"},{"key":"958_CR22","unstructured":"Li C, Li L, Jiang H et al (2022) YOLOv6: A single-stage object detection framework for industrial applications. arXiv preprint arXiv:2209.02976"},{"issue":"12","key":"958_CR23","doi-asserted-by":"publisher","first-page":"937","DOI":"10.3390\/insects15120937","volume":"15","author":"Y Li","year":"2024","unstructured":"Li Y, Chen Q, Zhu J, Li Z, Wang M, Zhang Y (2024) PM-YOLO: A powdery mildew automatic grading detection model for rubber tree. Insects 15(12):937","journal-title":"Insects"},{"key":"958_CR24","doi-asserted-by":"crossref","unstructured":"Li Y, Wang T, Luo N, Zhou L, Chen Q (2025a) CGMamba: Intelligent identification of counterfeit goods based on state space models. Int J Intell Syst 1:9939880","DOI":"10.1155\/int\/9939880"},{"key":"958_CR25","doi-asserted-by":"crossref","unstructured":"Li Y, Ye L, Zhang Y, Zhou L, Chen Q (2025b) Lightweight grading segmentation network for powdery mildew based on model pruning and knowledge distillation. Smart Agricultural Technology 101319","DOI":"10.1016\/j.atech.2025.101319"},{"issue":"11","key":"958_CR26","doi-asserted-by":"crossref","first-page":"925","DOI":"10.3390\/jof7110925","volume":"7","author":"Z Liu","year":"2021","unstructured":"Liu Z, Zhou Y, Chen X (2021) Detection of powdery mildew on rubber leaves based on hyperspectral imaging technology. J Fungi 7(11):925","journal-title":"J Fungi"},{"key":"958_CR27","doi-asserted-by":"crossref","unstructured":"Liu W, Anguelov D, Erhan D, Szegedy C, Reed S, Fu C-Y, Berg AC (2016) SSD: Single shot multibox detector. In: European conference on computer vision, pp 21\u201337. Springer","DOI":"10.1007\/978-3-319-46448-0_2"},{"key":"958_CR28","doi-asserted-by":"crossref","unstructured":"Liu Z, Li J, Shen Z, Huang G, Yan S, Zhang C (2017) Learning efficient convolutional networks through network slimming. In: Proceedings of the IEEE international conference on computer vision, pp 2736\u20132744","DOI":"10.1109\/ICCV.2017.298"},{"key":"958_CR29","doi-asserted-by":"crossref","unstructured":"Liu Y, Tian Y, Zhao Y, Yu H, Xie L, Wang Y, Shen Q, Liu Y (2024) VMamba: Visual state space model. arXiv preprint arXiv:2401.10166","DOI":"10.52202\/079017-3273"},{"key":"958_CR30","doi-asserted-by":"publisher","first-page":"378","DOI":"10.1016\/j.neucom.2017.06.023","volume":"267","author":"Y Lu","year":"2017","unstructured":"Lu Y, Yi S, Zeng N, Liu Y, Zhang Y (2017) Identification of rice diseases using deep convolutional neural networks. Neurocomputing 267:378\u2013384","journal-title":"Neurocomputing"},{"key":"958_CR31","doi-asserted-by":"crossref","unstructured":"Meng D, Chen X, Fan Z, Zeng G, Li H, Yuan Y, Sun L, Wang J (2021) Conditional DETR for fast training convergence. In: Proceedings of the IEEE\/CVF international conference on computer vision, pp 3651\u20133660","DOI":"10.1109\/ICCV48922.2021.00363"},{"key":"958_CR32","unstructured":"Ministry of Agriculture of the People\u2019s Republic of China (2015) NY\/T 1089\u20132015: Technical Procedure for Forecasting the Powdery Mildew of Rubber Tree. Agricultural Industry Standard of the People\u2019s Republic of China"},{"key":"958_CR33","doi-asserted-by":"publisher","first-page":"1419","DOI":"10.3389\/fpls.2016.01419","volume":"7","author":"SP Mohanty","year":"2016","unstructured":"Mohanty SP, Hughes DP, Salath\u00e9 M (2016) Using deep learning for image-based plant disease detection. Front Plant Sci 7:1419","journal-title":"Front Plant Sci"},{"issue":"1","key":"958_CR34","doi-asserted-by":"publisher","first-page":"103","DOI":"10.3390\/info17010103","volume":"17","author":"Z Qin","year":"2026","unstructured":"Qin Z, Li Y (2026) DCAM-DETR: Dual cross-attention mamba detection transformer for RGB-infrared anti-UAV detection. Information 17(1):103","journal-title":"Information"},{"key":"958_CR35","doi-asserted-by":"crossref","unstructured":"Qin D, Leichner C, Delakis M, Fornoni M, Luo S, Yang F, Wang W, Banbury C, Ye C, Akin B et al (2024) MobileNetV4 \u2013 universal models for the mobile ecosystem. arXiv preprint arXiv:2404.10518","DOI":"10.1007\/978-3-031-73661-2_5"},{"key":"958_CR36","doi-asserted-by":"crossref","unstructured":"Redmon J, Divvala S, Girshick R, Farhadi A (2016) You only look once: Unified, real-time object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 779\u2013788","DOI":"10.1109\/CVPR.2016.91"},{"key":"958_CR37","unstructured":"Ren S, He K, Girshick R, Sun J (2015) Faster R-CNN: Towards real-time object detection with region proposal networks. In: Adv Neural Inf Process Sys 28"},{"key":"958_CR38","doi-asserted-by":"crossref","unstructured":"Rezatofighi H, Tsoi N, Gwak J, Sadeghian A, Reid I, Savarese S (2019) Generalized intersection over union: A metric and a loss. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 658\u2013666","DOI":"10.1109\/CVPR.2019.00075"},{"key":"958_CR39","unstructured":"Romero A, Ballas N, Kahou SE, Chassang A, Gatta C, Bengio Y (2015) FitNets: Hints for thin deep nets. In: International Conference on Learning Representations"},{"key":"958_CR40","doi-asserted-by":"crossref","unstructured":"Sandler M, Howard A, Zhu M, Zhmoginov A, Chen L-C (2018) MobileNetV2: Inverted residuals and linear bottlenecks. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 4510\u20134520","DOI":"10.1109\/CVPR.2018.00474"},{"key":"958_CR41","unstructured":"Sun W, Jiang B, Liao E et al (2024) RT-DETRv2: Improved baseline with bag-of-freebies for real-time detection transformer. arXiv preprint arXiv:2407.17140"},{"issue":"1","key":"958_CR42","first-page":"1","volume":"20","author":"EJ Thomas","year":"2017","unstructured":"Thomas EJ, Molony B, Hoa TTT (2017) Rubber plantation and climate change: A review on improving rubber tree performance. J Rubb Res 20(1):1\u201330","journal-title":"J Rubb Res"},{"key":"958_CR43","doi-asserted-by":"crossref","unstructured":"Tian Z, Shen C, Chen H, He T (2019) FCOS: Fully convolutional one-stage object detection. In: Proceedings of the IEEE\/CVF international conference on computer vision, pp 9627\u20139636","DOI":"10.1109\/ICCV.2019.00972"},{"key":"958_CR44","doi-asserted-by":"crossref","unstructured":"Tian Y, Ye Q, Doermann D (2025) YOLOv12: Attention-centric real-time object detectors. arXiv preprint arXiv:2502.12524","DOI":"10.52202\/085713-2627"},{"issue":"8","key":"958_CR45","doi-asserted-by":"publisher","first-page":"555","DOI":"10.3390\/drones9080555","volume":"9","author":"S Umirzakova","year":"2025","unstructured":"Umirzakova S, Muksimova S, Buriboev AS, Primova H, Choi AJ (2025) A unified transformer model for simultaneous cotton boll detection, pest damage segmentation, and phenological stage classification from UAV imagery. Drones 9(8):555","journal-title":"Drones"},{"key":"958_CR46","doi-asserted-by":"crossref","unstructured":"Wang C-Y, Bochkovskiy A, Liao H-YM (2022) YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. arXiv preprint arXiv:2207.02696","DOI":"10.1109\/CVPR52729.2023.00721"},{"key":"958_CR47","doi-asserted-by":"crossref","unstructured":"Wang C, He W, Nie Y, Guo J, Liu C, Han K, Wang Y (2023) Gold-YOLO: Efficient object detector via gather-and-distribute mechanism. arXiv preprint arXiv:2309.11331","DOI":"10.52202\/075280-2224"},{"key":"958_CR48","doi-asserted-by":"crossref","unstructured":"Wang C-Y, Yeh I-H, Liao H-YM (2024a) YOLOv9: Learning what you want to learn using programmable gradient information. arXiv preprint arXiv:2402.13616","DOI":"10.1007\/978-3-031-72751-1_1"},{"key":"958_CR49","doi-asserted-by":"crossref","unstructured":"Wang A, Chen H, Liu L, Chen K, Lin Z, Han J, Ding G (2024b) YOLOv10: Real-time end-to-end object detection. arXiv preprint arXiv:2405.14458","DOI":"10.52202\/079017-3429"},{"key":"958_CR50","unstructured":"Wang M, Deng J, Chen Z et al (2025) YOLOv13: Real-time object detection with hypergraph-enhanced adaptive aggregation. arXiv preprint arXiv:2503.01267"},{"key":"958_CR51","unstructured":"Xu S, Wang X, Lv W, Chang Q, Cui C, Deng K et al (2022) PP-YOLOe: An evolved version of YOLO. arXiv preprint arXiv:2203.16250"},{"key":"958_CR52","doi-asserted-by":"crossref","unstructured":"Zhao B, Cui Q, Song R, Qiu Y, Liang J (2022) Decoupled knowledge distillation. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 11953\u201311962","DOI":"10.1109\/CVPR52688.2022.01165"},{"key":"958_CR53","doi-asserted-by":"crossref","unstructured":"Zhao Y, Lv W, Xu S, Wei J, Wang G, Dang Q, Liu Y, Chen J (2023) DETRs beat YOLOs on real-time object detection. arXiv preprint arXiv:2304.08069","DOI":"10.1109\/CVPR52733.2024.01605"},{"issue":"3","key":"958_CR54","doi-asserted-by":"publisher","first-page":"142","DOI":"10.3390\/technologies14030142","volume":"14","author":"D Zhou","year":"2026","unstructured":"Zhou D, Li Y (2026) SoccerDETR: Real-time soccer object detection via visual state space models with semantic-aware feature fusion. Technologies 14(3):142","journal-title":"Technologies"},{"key":"958_CR55","unstructured":"Zhu L, Liao B, Zhang Q, Wang X, Liu W, Wang X (2024) Vision mamba: Efficient visual representation learning with bidirectional state space model. arXiv preprint arXiv:2401.13260"},{"key":"958_CR56","doi-asserted-by":"crossref","unstructured":"Zong Z, Song G, Liu Y (2023) DETRS with collaborative hybrid assignments training, pp 6748\u20136758","DOI":"10.1109\/ICCV51070.2023.00621"}],"container-title":["Journal of King Saud University Computer and Information Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s44443-026-00958-2","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s44443-026-00958-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s44443-026-00958-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,11]],"date-time":"2026-08-11T11:10:58Z","timestamp":1786446658000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s44443-026-00958-2"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,30]]},"references-count":56,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2026,8]]}},"alternative-id":["958"],"URL":"https:\/\/doi.org\/10.1007\/s44443-026-00958-2","relation":{},"ISSN":["1319-1578","2213-1248"],"issn-type":[{"value":"1319-1578","type":"print"},{"value":"2213-1248","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,30]]},"assertion":[{"value":"19 April 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 June 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"30 June 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable.","order":2,"name":"Ethics","label":"Ethics approval and consent to participate","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"To facilitate reproducibility, we will release upon acceptance: (1) the complete training and evaluation source code based on PyTorch 2.1; (2) pretrained weights for both the full LightPM-DETR and compressed LightPM-DETR-C models; (3) training scripts with exact hyperparameters, random seeds (42, 123, 256), and data augmentation configurations; (4) TensorRT export scripts for edge deployment on NVIDIA Jetson platforms. All experiments were conducted on NVIDIA A100 GPUs with CUDA 12.1 and cuDNN 8.9. The exact software environment is specified in a\n requirements.txt\n file included in the repository.","order":3,"name":"Ethics","label":"Reproducibility Statement","group":{"name":"EthicsHeading","label":"Declarations"}}],"article-number":"565"}}