I have 2 datasets containing Leaf Images for Identification of Medicinal Plants of Bangladesh. These datasets are installed on Google Drive.

The first data set is the Original data set and it contains 10 class files. In total; There are 1440 Train images, 406 Validation images, 202 Test images.

The second data set is the Augmented data set and it contains 10 class files. In total; There are 27032 Train images, 7717 Validation images, 3857 Test images.

To obtain acceptable accuracy values, CNN architecture must be implemented.

Additionally, VGG and Inception architectural model should be implemented.

Accuracy values for CNN - VGG - Inception models should be compared with the table. Additionally, the Presicion - Recall and F1 results for each solution should be shown in the table.

Confusion Matrix, Training / Validation Loss and Training / Validation Accuracy values should be displayed.

Validation and Test results should be shown by commenting.