{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T06:18:10Z","timestamp":1787811490880,"version":"build-2784847793"},"reference-count":21,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T00:00:00Z","timestamp":1787788800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Plant Sci."],"abstract":"<jats:p>\n                    Pomegranate cultivation faces significant productivity losses due to fungal and bacterial diseases, yet existing automated detection systems rely on centralized deep learning pipelines that require raw image data aggregation, raising data privacy concerns in distributed agricultural settings. This paper proposes a federated learning (FL) framework for raw-data-decentralized pomegranate disease classification, systematically evaluating six architecturally diverse deep learning models such as Custom CNN, ResNet50, ConvNeXt_V2, ViT-B\/16, EfficientNetV2-S, and MobileViT-S (under the FedAvg aggregation protocol across five simulated clients and 10 communication rounds). All six architectures are initialized with publicly available ImageNetpretrained weights to ensure a fair architectural comparison. Experiments are conducted on a working set of 6,823 images derived from the Mendeley Pomegranate Fruit Diseases Dataset (5,099 original images) spanning five disease classes\u2014\n                    <jats:italic>Alternaria<\/jats:italic>\n                    , anthracnose, bacterial blight,\n                    <jats:italic>Cercospora<\/jats:italic>\n                    fruit spot, and healthy\u2014using a split-before-augmentation pipeline: the train\/validation\/test split is performed prior to any augmentation; augmentation is applied only to the training partition (5,800 images), while the validation (511) and test (512) sets contain only original, unaugmented images. EfficientNetV2-S achieves the single highest test accuracy of 99.22%, with ViT-B\/16, ResNet50, and ConvNeXt_V2 tied at 98.83% and MobileViT-S close behind at 98.63%, offering the strongest efficiency\u2013accuracy tradeoff. We further report results under non-IID (Dirichlet-partitioned) client distributions, centralized and local-only training baselines, and multi-seed statistical evaluation. All six models are exported to ONNX format and deployed as a real-time web application on Hugging Face Spaces, demonstrating practical deployment readiness. The results confirm the viability of federated learning for high-accuracy, raw-data-decentralized crop disease classification while also characterizing its performance degradation under realistic non-IID conditions.\n                  <\/jats:p>","DOI":"10.3389\/fpls.2026.1906269","type":"journal-article","created":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T05:34:04Z","timestamp":1787808844000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["FedPome: federated deep learning for real-time pomegranate disease classification"],"prefix":"10.3389","volume":"17","author":[{"given":"Lokesh","family":"S","sequence":"first","affiliation":[{"name":"School of Computer Science Engineering and Information Systems, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Akshaya Prathiksha","family":"C. K.","sequence":"additional","affiliation":[{"name":"School of Electronics Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aishwaryalakshmi","family":"M.","sequence":"additional","affiliation":[{"name":"School of Electronics Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vishalini","family":"V.","sequence":"additional","affiliation":[{"name":"School of Electronics Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yenugonda Hari","family":"Krishna","sequence":"additional","affiliation":[{"name":"School of Computer Science Engineering and Information Systems, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aparna","family":"Mohanty","sequence":"additional","affiliation":[{"name":"School of Electronics Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,27]]},"reference":[{"key":"B1","first-page":"712","article-title":"Disease detection and classification in pomegranate fruit using hybrid convolutional neural network with honey badger optimization algorithm","volume":"130","author":"Bhange","year":"2023","journal-title":"J. Plant Dis. Prot."},{"key":"B2","doi-asserted-by":"publisher","first-page":"280","DOI":"10.1016\/j.procs.2015.08.022","article-title":"Smart farming: Pomegranate disease detection using image processing","volume":"58","author":"Bhange","year":"2015","journal-title":"Proc. Comput. Sci."},{"key":"B3","article-title":"An image is worth 16\u00d716 words: Transformers for image recognition at scale","author":"Dosovitskiy","year":"2021"},{"key":"B4","first-page":"770","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B5","doi-asserted-by":"publisher","first-page":"112382","DOI":"10.1016\/j.scienta.2023.112382","article-title":"A deep learning approach to detect diseases in pomegranate fruits via hybrid optimal attention capsule network","volume":"322","author":"Kumar","year":"2023","journal-title":"Sci. Hortic."},{"key":"B6","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"LeCun","year":"1998","journal-title":"Proc. IEEE"},{"key":"B7","doi-asserted-by":"publisher","first-page":"50","DOI":"10.1109\/msp.2020.2975749","article-title":"Federated learning: Challenges, methods, and future directions","volume":"37","author":"Li","year":"2020","journal-title":"IEEE Signal Process Mag."},{"key":"B8","first-page":"11976","article-title":"A Convnet for the 2020s","author":"Liu","year":"2022"},{"key":"B9","first-page":"1273","article-title":"Communication-efficient learning of deep networks from decentralized data","author":"McMahan","year":"2017"},{"key":"B10","article-title":"MobileViT: Light-weight, general-purpose, and mobile-friendly vision transformer","author":"Mehta","year":"2022"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.1016\/j.dib.2025.112298","article-title":"Pomegranate disease detection and classification dataset for deep learning applications: A case study from Halabja city","volume":"58","author":"Mohammed","year":"2025","journal-title":"Data Brief"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.3389\/fpls.2016.01419","article-title":"Using deep learning for image-based plant disease detection","volume":"7","author":"Mohanty","year":"2016","journal-title":"Front. Plant Sci."},{"key":"B13","doi-asserted-by":"publisher","DOI":"10.17632\/b6s2rkpmvh.1","article-title":"Pomegranate fruit diseases dataset for deep learning models","author":"Pakruddin","year":"2023","journal-title":"Mendeley Data"},{"key":"B14","doi-asserted-by":"publisher","first-page":"1535","DOI":"10.18805\/ijare.a-6396","article-title":"Performance evaluation of deep learning models for multiclass disease detection in pomegranate fruits","volume":"59","author":"Pakruddin","year":"2025","journal-title":"Indian J. Agric. Res."},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.3389\/fpls.2017.01852","article-title":"Deep learning for image-based cassava disease detection","volume":"8","author":"Ramcharan","year":"2017","journal-title":"Front. Plant Sci."},{"key":"B16","doi-asserted-by":"publisher","first-page":"41278","DOI":"10.1038\/s41598-025-25160-8","article-title":"Pomegranate disease diagnosis with severity estimation and treatment remedies using deep learning and RAG-based LLM","volume":"15","author":"Revathi","year":"2025","journal-title":"Sci. Rep."},{"key":"B17","doi-asserted-by":"publisher","first-page":"124","DOI":"10.1186\/s13007-025-01447-9","article-title":"Optimized deep learning framework for pomegranate disease detection using natureinspired algorithms","volume":"21","author":"Sandhi","year":"2025","journal-title":"Plant Methods"},{"key":"B18","first-page":"618","article-title":"Grad-CAM: Visual explanations from deep networks via gradient-based localization","author":"Selvaraju","year":"2017"},{"key":"B19","first-page":"10096","article-title":"EfficientNetV2: Smaller models and faster training","author":"Tan","year":"2021"},{"key":"B20","article-title":"PyTorch image models (timm)","volume-title":"Github","author":"Wightman","year":"2019"},{"key":"B21","first-page":"16133","article-title":"ConvNeXt V2: Co-designing and scaling ConvNets with masked autoencoders","author":"Woo","year":"2023"}],"container-title":["Frontiers in Plant Science"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fpls.2026.1906269\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T05:34:05Z","timestamp":1787808845000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fpls.2026.1906269\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,27]]},"references-count":21,"alternative-id":["10.3389\/fpls.2026.1906269"],"URL":"https:\/\/doi.org\/10.3389\/fpls.2026.1906269","relation":{},"ISSN":["1664-462X"],"issn-type":[{"value":"1664-462X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,27]]},"article-number":"1906269"}}