farmer-nusantara / farmer-nusantara-machine-learning

Deep Learning Model for Classification plant handlers

farmer-nusantara-machine-learning

Overview :
For this stage, machine learning has built 3 kinds of image classification models, one model is representative of one food plant, there are Paddy, Corn, and Soy Bean. We use a public dataset from the Kaggle platform for that classification. Paddy has 10 classes, corn 4 classes, and 3 classes for soybean. Those 3 datasets include the health plant conditions class

Dataset Link :

• Paddy Leaves
  • Paddy Doctor: Paddy Disease Classification
• Corn Leaves → Modified Dataset
  • Plant Village Dataset
  • Corn Leaves Dataset
• Soybean Leaves
  • Soybean Leaves Dataset

Paddy Model Best Accuracy :
Train Accuracy = 0.9828
Validation Accuracy = 0.9521

Corn Model Best Accuracy :
Train Accuracy = 0.9828
Validation Accuracy = 0.9521

Soybean Model Best Accuracy :
Train Accuracy = 0.9977
Validation Accuracy = 0.9717

Prerequisites to Replicate

1. Google Colab
2. Kaggle API Token → Generate
3. Latest version of Tensorflow 2.8.2 (or you can update again by rerunning .ipynb and updating models)