Impact of Hyperparameter Tuning on Deep Learning Based Estimation of Disease Severity in Grape Plant
Accurate and quantitative estimation of disease severity in plants is a complex task, even for experienced agronomists and plant pathologists, where incorrect evaluation might lead to the inappropriate use of pesticides. In this paper, the authors have utilized two Convolutional Neural Networks (CNN), namely, AlexNet and ResNet18, for assessing the disease severity in Grape plant. The images for Isariopsis Leaf Spot disease in Grape plant, also known as Leaf Blight, were taken from the PlantVillage dataset, divided into three categories of severity stages (early, middle & end) and used for training the CNN models, via the transfer learning approach. The effects of fine-tuning the hyperparameters such as mini-batch size, epochs and data augmentation were also observed and analysed. For performance comparison, measures such as classification accuracy, mean F1-score, mean precision, mean recall, validation loss, ROC curves and time taken were recorded.
KeywordsDeep learning Convolutional neural networks (CNN) Hyperparameter tuning Plant diseases Disease severity Grape plant
This study was supported by Department of Science and Technology (DST), Government of India, New Delhi, under Interdisciplinary Cyber Physical Systems (ICPS) Programme (Reference No. T-319).
- 7.Mohanty SP, Hughes DP, Salathé M (2016) Using Deep Learning for Image-Based Plant Disease Detection. Frontiers in Plant Science. 7(September):1–10Google Scholar
- 9.Wang, G., Sun, Y., & Wang, J.: Automatic Image-Based Plant Disease Severity Estimation Using Deep Learning. Computational Intelligence and Neuroscience. 1–8, (2017)Google Scholar
- 12.Verma, S., Chug, A., Singh, A. P., Sharma, S., Rajvanshi, P.: Deep Learning-Based Mobile Application for Plant Disease Diagnosis: A Proof of Concept With a Case Study on Tomato Plant. In Applications of Image Processing and Soft Computing Systems in Agriculture. 242–271, IGI Global, (2019)Google Scholar
- 14.Sun Y, Liu Y, Wang G, Zhang H (2017) Deep learning for plant identification in natural environment. Comput Intell Neurosci 2017:1–6Google Scholar
- 17.Krizhevsky, A., Sutskever, I., Hinton, G. E.: Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems. 1097–1105, (2012)Google Scholar
- 18.He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition. 770–778, (2016)Google Scholar
- 19.Hughes, D. P., Salathe, M.: An open access repository of images on plant health to enable the development of mobile disease diagnostics. ArXiv Preprint. https://doi.org/10.1111/1755-0998.12237, (2015)
- 21.https://www.mathworks.com/help/deeplearning/ug/pretrained-convolutional-neural-networks.html. Last accessed: May 2019