Install the required libraries. Assuming the dataset is uploaded to Google Drive, pull the dataset from there. There are 8 different plant data folder in the data set and they are divided into 3 folders called test train validation.

To obtain acceptable accuracy values, we first need to create a CNN.

We need to apply at least two more CNN-based deep learning models such as VGG, Resnet, Inception, DenseNet architectures and compare the accuracy values of all these networks on a table. Let's also draw a table of precision, recall and F1 results for each class for each solution.

Represent the confusion matrix, training/validation loss, and training/validation accuracy values. Show verification and testing results by commenting.