Enhancing Rose Leaf Disease Detection Accuracy Using Optimized CNN Parameters

Latifah Nabilah; Nisar Nisar; Amnah Amnah; Septilia  Arfida

doi:10.25008/bcsee.v4i2.1184

Authors

Latifah Nabilah Department of Informatics Technology, Institute Informatics & Business Darmajaya
Nisar Nisar Department of Informatics Technology, Institute Informatics & Business Darmajaya
Amnah Amnah Department of Informatics Technology, Institute Informatics & Business Darmajaya
Septilia Arfida Department of Informatics Technology, Institute Informatics & Business Darmajaya

DOI:

https://doi.org/10.25008/bcsee.v4i2.1184

Keywords:

CNN model, Rose leaf disease detection, Optimization, RMSProp optimizer, Early stopping

Abstract

The CNN model developed in this study demonstrated remarkable performance, achieving an outstanding validation accuracy of 99.96%. Through experimentation, it was found that employing the RMSprop optimizer with a learning rate of 0.001 yielded superior results compared to the Adam optimizer utilized in previous iterations. Additionally, increasing the number of epochs from 10 to 20 resulted in a significant enhancement in accuracy, highlighting the importance of iterative training for model refinement. Moreover, the implementation of Early Stopping proved to be a valuable technique, effectively conserving training time by halting the training process once optimal accuracy levels were reached. These findings underscore the efficacy of various optimization strategies in bolstering the performance of CNN models for rose leaf disease detection. The achieved accuracy rates signify a substantial advancement in disease detection technology, holding promise for enhancing agricultural productivity and ensuring plant quality. This research contributes valuable insights into the optimization of CNN parameters, paving the way for further advancements in automated disease detection systems in the field of agriculture.

Downloads

Download data is not yet available.

References

S. Nuanmeesri, “A hybrid deep learning and optimized machine learning approach for rose leaf disease classification,” Eng. Technol. &Applied Sci. Res., 2021, [Online]. Available: https://www.etasr.com/index.php/ETASR/article/view/4455

“Produksi Tanaman Florikultura (Hias).”

M. Ali-Al-Alvy, G. K. Khan, M. J. Alam, S. Islam, M. Rahman, and M. S. Rahman, “Rose Plant Disease Detection using Deep Learning,” 7th Int. Conf. Trends Electron. Informatics, ICOEI 2023 - Proc., no. April, pp. 1244–1249, 2023, doi: 10.1109/ICOEI56765.2023.10126031.

S. K. Basak, “Unveiling the Enigma?: Advancing Rose Leaf Disease Detection with Transformed Images and Convolutional Neural Networks,” Zenodo, vol. 1, pp. 1–7, 2023.

A. Rajbongshi, T. Sarker, M. M. Ahamad, and M. M. Rahman, “Rose Diseases Recognition using MobileNet,” 4th Int. Symp. Multidiscip. Stud. Innov. Technol. ISMSIT 2020 - Proc., 2020, doi: 10.1109/ISMSIT50672.2020.9254420.

et al., “Plant Leaf Disease Detection and Classification Using Deep Learning: A Review and A Proposed System on Bangladesh’s Perspective,” Int. J. Sci. Bus., vol. 28, no. 1, pp. 193–204, 2023, doi: 10.58970/ijsb.2214.

Xiaoguang Lu, Yunhong Wang, and A. K. Jain, “Combining classifiers for face recognition,” no. March, pp. III–13, 2004, doi: 10.1109/icme.2003.1221236.

A. Mujhid, S. Surono, N. Irsalinda, and A. Thobirin, “Comparison and Combination of Leaky ReLU and ReLU Activation Function and Three Optimizers on Deep CNN for COVID-19 Detection,” Front. Artif. Intell. Appl., vol. 358, pp. 50–57, 2022, doi: 10.3233/FAIA220369.

Y. Bai, “RELU-Function and Derived Function Review,” SHS Web Conf., vol. 144, p. 02006, 2022, doi: 10.1051/shsconf/202214402006.

F. Schilling, “The Effect of Batch Normalization on Deep Convolutional Neural Networks,” p. 113, 2016, [Online]. Available: http://kth.diva-portal.org/smash/record.jsf?pid=diva2%3A955562&dswid=-5716

D. Bertoin, J. Bolte, S. Gerchinovitz, and E. Pauwels, “Numerical influence of ReLU’(0) on backpropagation,” Adv. Neural Inf. Process. Syst., vol. 1, no. 0, pp. 468–479, 2021.

P. Meißner, H. Watschke, J. Winter, and T. Vietor, “Artificial neural networks-based material parameter identification for numerical simulations of additively manufactured parts by material extrusion,” Polymers (Basel)., vol. 12, no. 12, pp. 1–28, 2020, doi: 10.3390/polym12122949.

T. Nagaoka, “Hyperparameter optimization for deep learning-based automatic melanoma diagnosis system,” Adv. Biomed. Eng., vol. 9, pp. 225–232, 2020, doi: 10.14326/abe.9.225.

K. Nakamura, B. Derbel, K. J. Won, and B. W. Hong, “Learning-rate annealing methods for deep neural networks,” Electron., vol. 10, no. 16, pp. 1–12, 2021, doi: 10.3390/electronics10162029.

M. Alalhareth and S. C. Hong, “An Adaptive Intrusion Detection System in the Internet of Medical Things Using Fuzzy-Based Learning,” Sensors (Basel)., vol. 23, no. 22, 2023, doi: 10.3390/s23229247.

T. Miseta, A. Fodor, and Á. Vathy-Fogarassy, “Surpassing early stopping: A novel correlation-based stopping criterion for neural networks,” Neurocomputing, vol. 567, no. November 2023, p. 127028, 2024, doi: 10.1016/j.neucom.2023.127028.

S. F. Ahmed et al., Deep learning modelling techniques: current progress, applications, advantages, and challenges, vol. 56, no. 11. Springer Netherlands, 2023. doi: 10.1007/s10462-023-10466-8.