Performance Results of Plant Disease Classification Pretrained Models
Description
This dataset contains the performance results of five pre-trained deep learning models—EfficientNetV2B0, ResNet50, InceptionV3, DenseNet121, and VGG16—evaluated for plant disease classification tasks. The dataset includes metrics from experiments conducted on both training and testing datasets, focusing on key performance indicators such as accuracy, precision, recall, F1-score, and confusion matrices. The aim is to provide a comprehensive comparison of these models' capabilities for detecting and classifying plant diseases.
Files
Steps to reproduce
To reproduce the results, prepare a labeled plant disease dataset (e.g., PlantVillage), split it into training, validation, and testing sets, and preprocess the images. Set up the environment with Python 3.8+ and required libraries (e.g., TensorFlow, Keras, scikit-learn). Load the pre-trained models (EfficientNetV2B0, ResNet50, InceptionV3, DenseNet121, VGG16), apply transfer learning by freezing base layers and adding custom classification layers, and compile using an optimizer (e.g., Adam) and loss function (e.g., categorical crossentropy). Train each model with callbacks like early stopping, evaluate on the testing set, and compute metrics such as accuracy, precision, recall, F1-score, and confusion matrices. Visualize results with graphs and confusion matrices, and download supplementary data and scripts from the Mendeley Dataset link to replicate the full pipeline.