Deep Learning Framework for Cashew Pest Classification and Automated Treatment Recommendation
International Journal of Innovative Research in Science Engineering and Technology · 24.07.2026 · 15 (7)
DOI: 10.15680/ijirset.2026.1507046 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Innovative Research in Science Engineering and Technology, online erschienen am 24.07.2026. Verfasst von Rythum Rythum und 1 weiteren Autor:innen.
Abstract (Originalsprache)
The rapid advancement of automated farm surveillance has created the cashew pests and diseases classification as a basis of precision farming. This is a literature review of the top ten most recent studies, that is, the methodologies and results that have been achieved in the field of literature related to transferring the traditional statistical learning into high-performing deep learning (DL) systems. The latest advances have significantly improved classification accuracy with the aid of convolutional neural networks (CNNs) as well as attention based on hybrid concatenated, vision transformer, and optimized recurrent neural networks. Although these approaches have shown outstanding performance on standard datasets like CCMT and CCDDB, issues of model interpretability, real-time edgedevice deployment and stability in various environmental field settings continue to be a problem. The paper puts the state-of-the-art methods into the categories depending on their architecture, including traditional SVM, Random Forest, against the latest EfficientNetV2-S and DenseNet. A comparative analysis of performance measures is provided, with an emphasis on the new tendencies related to hybrid CNNTransformer models, multi-scale feature fusion, and gradientweighted class activation mapping (Grad-CAM) as a visual elucidation method. The paper concludes with the identification of open research challenges and directions on how to create more scalable and generalizable cashew pest classification systems.
Zitieren
APA
Rythum, R., & Dhiman, N. (2026). Deep Learning Framework for Cashew Pest Classification and Automated Treatment Recommendation. International Journal of Innovative Research in Science Engineering and Technology, 15(7). https://doi.org/10.15680/ijirset.2026.1507046
Vancouver
Rythum R, Dhiman N. Deep Learning Framework for Cashew Pest Classification and Automated Treatment Recommendation. International Journal of Innovative Research in Science Engineering and Technology. 2026;15(7). doi:10.15680/ijirset.2026.1507046
BibTeX
@article{Rythum2026,
author = {Rythum, Rythum and Dhiman, Neha},
title = {Deep Learning Framework for Cashew Pest Classification and Automated Treatment Recommendation},
journal = {International Journal of Innovative Research in Science Engineering and Technology},
year = {2026},
volume = {15},
number = {7},
doi = {10.15680/ijirset.2026.1507046}
}
EUVIMED