Fig 1.
Images of maize leaf illness sample.
Table 1.
Chronicle of abbreviations.
Table 2.
Comparative analysis of various works pertaining to crop disease detection and classification.
Fig 2.
Proposed architecture for maize leaf disease classification.
Table 3.
Sample of dataset.
Table 4.
Dataset utilized.
Fig 3.
14-Layer convolutional neural network.
Fig 4.
Deep transfer learning with Bayesian hyper-parameter optimization.
Fig 5.
Flow diagram for feature selection with HHOSA.
Table 5.
Fuzzy rule algorithm for severity assessment.
Table 6.
Description of severity level.
Table 7.
Hyper-parameters utilized for our proposed work.
Fig 6.
Training vs validation accuracy.
Fig 7.
Training vs validation loss.
Table 8.
Prediction accuracy and misclassification rate.
Fig 8.
Confusion matrix for the proposed methodology.
Fig 9.
Provides the prediction performance in terms of a confusion matrix based on ResNet5.
Fig 10.
Provides the prediction performance in terms of confusion matrix based on ResNet152.
Fig 11.
Provides the prediction performance in terms of the confusion matrix based on InceptionV3.
Fig 12.
Provides the prediction performance in terms of the confusion matrix based on DenseNet152.
Fig 13.
Provides the prediction performance in terms of the confusion matrix for the proposed model based on EfficientNet-B7.
Table 9.
Comparative analysis of class-wise prediction performance.
Table 10.
Comparative analysis of transfer learning models.
Fig 14.
Comparative analysis of performance metrics of the proposed and the deep transfer learning models in terms of classification accuracy.
Fig 15.
Comparative analysis of performance metrics between the proposed models and deep transfer learning models in terms of testing loss.
Fig 16.
Comparative analysis of performance metrics of the proposed and the deep transfer learning models in terms of average precision and sensitivity.
Fig 17.
Comparative analysis of performance metrics of the proposed models and deep transfer learning models in terms of average F1-score.
Table 11.
Comparative analysis of prediction performance to that of existing systems.
Table 12.
Comparative analysis of accuracy for different image augmentation techniques.
Table 13.
Comparison of performance of Existing works and Proposed solution with different data volumes.