SEC-ToLeD: A Super-Efficient and Compact CNN for Tomato Leaves Diseases Detection for Resource-Constrained Devices

Authors

  • Nissreen El-Saber Department of Information Systems, Faculty of Computers and Informatics, Zagazig University, Zagazig 44519, Egypt, Department of Software Engineering, Faculty of Information and Computers, Misr International University, Obour City, Qalyubiyya Governorate, Egypt https://orcid.org/0000-0002-3988-1437
  • ‪Fady Salama‬‏ Department of Information Systems, Faculty of Computers and Informatics, Zagazig University, Zagazig 44519, Egypt, Department of Information Systems, Faculty of Computers and Information Technology, Innovation University, 10th of Ramadan City, Egypt.
  • Islam Samy Department of Information Systems, Faculty of Computers and Informatics, Zagazig University, Egypt.

DOI:

https://doi.org/10.19139/soic-2310-5070-3579

Keywords:

Convolutional neural networks, soft attention, deep learning, tomato, plant disease detection, squeeze and-excitation (SE) blocks

Abstract

Tomato leaf diseases significantly affect agricultural productivity and require accurate yet lightweightdetection models suitable for real-world deployment. In this study, a super-efficient and compact convolutionalneural network, named SEC-ToLeD, is proposed for tomato leaf disease classification using the PlantVillagedataset. The model integrates depthwise separable convolutions with squeeze-and-excitation blocks to enhancefeature representation while maintaining low computational complexity. SEC-ToLeD contains only 219,651trainable parameters with a compact size of 0.84 MB. Experimental results demonstrate 99.75% testing accuracy, 99% precision, recall, and F1-score, with an average inference time of 1.105 ms per image. Compared to existing lightweight models, SEC-ToLeD achieves superior performance while being significantly smaller in size, making it highly suitable for real-time deployment on resource-constrained and edge devices.

Downloads

Published

2026-04-16

Issue

Online First

Section

Research Articles

License

Authors who publish with this journal agree to the following terms:

    1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
    2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
    3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).

How to Cite

SEC-ToLeD: A Super-Efficient and Compact CNN for Tomato Leaves Diseases Detection for Resource-Constrained Devices. (2026). Statistics, Optimization & Information Computing. https://doi.org/10.19139/soic-2310-5070-3579