Original Articles Open access

Efficient and explainable attention-based multitask learning for fruit type, ripeness, and disease classification

Authors

Published
25 September 2026
Issue
Early Access
Section
Original Articles
DOI
10.4081/jae.2026.2108

Abstract

Precision agriculture increasingly relies on automated crop monitoring to support disease prevention and harvest decisions. However, most existing systems treat fruit type, ripeness, and disease classification as separate tasks, leading to high computational costs in real-world deployments. This paper proposes an efficient and explainable multitask learning (MTL) model for fruit monitoring that performs all three tasks simultaneously. The MTL model uses a shared lightweight MobileNetV2 backbone with a convolutional block attention module (CBAM) to extract both shared and discriminative features. Dedicated classification heads then utilize these features to generate predictions for each task. To address the scarcity of fully annotated agricultural datasets, we combine multiple public datasets and apply a partial label-masking strategy to handle missing annotations. In addition, the proposed model is evaluated using several backbone architectures, including ResNet-50, EfficientNetV2-M, and ViT-B/16, under both single optimizer and separate optimizers. Experimental results demonstrate that the proposed MTL model achieves 99.39% (fruit type), 96.35% (ripeness), and 99.7% (disease) accuracy, while using only 5.5M parameters, outperforming single-task solutions. Furthermore, gradient-based explainability analysis provides transparency into the model's decision-making process.

Keywords

Data Availability Statement

All the data and materials used are included and referenced in this paper. The source code and implementation details are publicly available at: https://github.com/ml-research-ul/AgriMTL

References

  1. Ahmed MS, 2023. Fruit Image Dataset: 22 Classes. Kaggle. Available from: https://www.kaggle.com/datasets/mdsagorahmed/fruit-image-dataset-22-classes
  2. Ali A, 2019. PlantVillage Dataset. Kaggle. Available from: https://www.kaggle.com/datasets/abdallahalidev/plantvillage-dataset
  3. Atila Ü, Uçar M, Akyol K, Uçar E, 2021. Plant leaf disease classification using EfficientNet deep learning model. Ecol Inform 61:101182. DOI: https://doi.org/10.1016/j.ecoinf.2020.101182
  4. Attri I, Awasthi LK, Sharma TP, Rathee P, 2023. A review of deep learning techniques used in agriculture. Ecol Inform 77:102217. DOI: https://doi.org/10.1016/j.ecoinf.2023.102217
  5. Bai C, Zhang L, Gao L, Peng L, Li P, Yang L, 2024. DINOV2-FCS: a model for fruit leaf disease classification and severity prediction. Front Plant Sci 15:1475282. DOI: https://doi.org/10.3389/fpls.2024.1475282
  6. Bhattarai S, 2018. New Plant Diseases Dataset. Kaggle. Available from: https://www.kaggle.com/datasets/vipoooool/new-plant-diseases-dataset
  7. Dai G, Tian Z, Fan J, Sunil CK, Dewi C, 2024. DFN-PSAN: multi-level deep information feature fusion extraction network for interpretable plant disease classification. Comput Electron Agric 216:108481. DOI: https://doi.org/10.1016/j.compag.2023.108481
  8. Diyansah D, 2024. Fruit Ripeness Level Dataset. Kaggle. Available from: https://www.kaggle.com/datasets/dudinurdiyansah/fruit-ripeness-level-dataset/data
  9. Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner T, et al., 2021. An image is worth 16×16 words: transformers for image recognition at scale. arXiv:2010.11929.
  10. Duarte K, Rawat YS, Shah M, 2021. PLM: partial label masking for imbalanced multi-label classification. arXiv:2105.10782. DOI: https://doi.org/10.1109/CVPRW53098.2021.00308
  11. Fu R, Wang X, Wang S, Sun H, 2025. PMJDM: a multi-task joint detection model for plant disease identification. Front Plant Sci 16:1599671. DOI: https://doi.org/10.3389/fpls.2025.1599671
  12. Hamed BS, Hussein MM, Mousa AM, 2023. Plant disease detection using deep learning. Int J Intell Syst Appl 15:38-50. DOI: https://doi.org/10.5815/ijisa.2023.06.04
  13. He K, Zhang X, Ren S, Sun J, 2015. Deep residual learning for image recognition. arXiv:1512.03385. DOI: https://doi.org/10.1109/CVPR.2016.90
  14. Hemalatha S, Jayachandran JJB, 2024. A multitask learning-based vision transformer for plant disease localization and classification. Int J Comput Intell Syst 17:188. DOI: https://doi.org/10.1007/s44196-024-00597-3
  15. Hu D, Zhao L, Ren Y, Wang S, Ye X, Zhang H, et al., 2025. A lightweight vision transformer with embedded hybrid attention for quick response code defect classification. Eng Appl Artif Intell 150:110575. DOI: https://doi.org/10.1016/j.engappai.2025.110575
  16. Krishna MS, Machado P, Otuka RI, Yahaya SW, Neves dos Santos F, Ihianle IK, 2025. Plant leaf disease detection using deep learning: a multi-dataset approach. J 8:4. DOI: https://doi.org/10.3390/j8010004
  17. Mi Z, Yan WQ, 2024. Strawberry ripeness detection using deep learning models. Big Data Cogn Comput 8:92. DOI: https://doi.org/10.3390/bdcc8080092
  18. Nair S, Bajpai K, Fadnavis S, Jain S, Sheikh Z, 2024. Plant Fettle Detector: unifying fruit quality tracking and plant disease detection through advanced image analysis. In: Proceedings of the 2024 3rd International Conference for Innovation in Technology (INOCON). IEEE; pp. 1-5. DOI: https://doi.org/10.1109/INOCON60754.2024.10512262
  19. Padhiary M, Saha D, Kumar R, Sethi LN, Kumar A, 2024. Enhancing precision agriculture: a comprehensive review of machine learning and AI vision applications in all-terrain vehicles for farm automation. Smart Agric Technol 8:100483. DOI: https://doi.org/10.1016/j.atech.2024.100483
  20. Sandhi A, 2024. A comprehensive hybrid model for apple fruit disease detection using multi-architecture feature extraction. Preprints 202410.1894.v1. DOI: https://doi.org/10.20944/preprints202410.1894.v1
  21. Sandler M, Howard A, Zhu M, Zhmoginov A, Chen LC, 2019. MobileNetV2: inverted residuals and linear bottlenecks. arXiv:1801.04381. DOI: https://doi.org/10.1109/CVPR.2018.00474
  22. Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, Batra D, 2017. Grad-CAM: visual explanations from deep networks via gradient-based localization. In: Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV). IEEE; pp. 618-26. DOI: https://doi.org/10.1109/ICCV.2017.74
  23. Singh D, Jain N, Jain P, Kayal P, Kumawat S, Batra N, 2020. PlantDoc: a dataset for visual plant disease detection. In: Proceedings of the 7th ACM IKDD CoDS and 25th COMAD (CoDS COMAD 2020). Association for Computing Machinery; pp. 249-53. DOI: https://doi.org/10.1145/3371158.3371196
  24. sshikamaru, 2018. Fruit Recognition. Kaggle. Available from: https://www.kaggle.com/datasets/sshikamaru/fruit-recognition
  25. Sultana N, Jahan M, Uddin MS, 2022. An extensive dataset for successful recognition of fresh and rotten fruits. Data Brief 44:108552. DOI: https://doi.org/10.1016/j.dib.2022.108552
  26. Tan M, Le QV, 2021. EfficientNetV2: smaller models and faster training. arXiv:2104.00298.
  27. Tapia-Mendez E, Cruz-Albarran IA, Tovar-Arriaga S, Morales-Hernandez LA, 2023. Deep learning-based method for classification and ripeness assessment of fruits and vegetables. Appl Sci 13:12504. DOI: https://doi.org/10.3390/app132212504
  28. Uddin AH, 2020. Plant Doc Dataset. Kaggle. Available from: https://www.kaggle.com/datasets/abdulhasibuddin/plant-doc-dataset
  29. Wang CY, Yeh IH, Liao HYM, 2024. YOLOv9: learning what you want to learn using programmable gradient information. arXiv:2402.13616. DOI: https://doi.org/10.1007/978-3-031-72751-1_1
  30. Wang P, Niu T, Mao Y, Zhang Z, Liu B, He D, 2021. Identification of apple leaf diseases by improved deep convolutional neural networks with an attention mechanism. Front Plant Sci 12:723294. DOI: https://doi.org/10.3389/fpls.2021.723294
  31. Woo S, Park J, Lee JY, Kweon IS, 2018. CBAM: convolutional block attention module. arXiv:1807.06521. DOI: https://doi.org/10.1007/978-3-030-01234-2_1
  32. Yang B, Wang Z, Guo J, Guo L, Liang Q, Zeng Q, et al., 2023. Identifying plant disease and severity from leaves: a deep multitask learning framework using triple-branch Swin Transformer and deep supervision. Comput Electron Agric 209:107809. DOI: https://doi.org/10.1016/j.compag.2023.107809
  33. Yu J, Dai Y, Liu X, Huang J, Shen Y, Zhang K, et al., 2024. Unleashing the power of multi-task learning: a comprehensive survey spanning traditional, deep, and pretrained foundation model eras. arXiv:2404.18961.
  34. Zhang W, 2023. A fruit ripeness detection method using adapted deep learning-based approach. Int J Adv Comput Sci Appl 14:121. DOI: https://doi.org/10.14569/IJACSA.2023.01409121
  35. Zhang Y, Yang X, Cheng Y, Wu X, Sun X, Hou R, et al., 2024. Fruit freshness detection based on multi-task convolutional neural network. Curr Res Food Sci 8:100733. DOI: https://doi.org/10.1016/j.crfs.2024.100733

Metrics

Citations and online attention from Dimensions, Altmetric and PlumX.

PlumX Metrics

License

Copyright (c) 2026 the Author(s)

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

PAGEPress has chosen to apply the Creative Commons Attribution NonCommercial 4.0 International License (CC BY-NC 4.0) to all manuscripts to be published.