AAAkater/VGG-CNN

VGG16-CNN Rice Disease Classification Model

This model is designed for classifying rice plant diseases using a modified VGG16 architecture with additional CNN layers.

Model Description

Architecture

• Base model: VGG16 (pretrained on ImageNet)
• Additional custom CNN layer with:
  • Conv2d(512, 64, kernel_size=3)
  • ReLU activation
  • BatchNorm2d
  • MaxPool2d
• Custom classifier with:
  • Linear layers (3236 → 1024 → 5)
  • Dropout (0.4)

Task

Image classification for rice plant diseases

Classes

1. Bacterialblight
2. Blast
3. Brownspot
4. Healthy
5. Tungro

Training

The model uses transfer learning with a frozen VGG16 backbone.

Intended Use

• Primary intended use: Rice disease diagnosis through leaf image analysis
• Out-of-scope use: Should not be used for critical agricultural decisions without expert verification

Input

• RGB images
• Required size: 224x224 pixels
• Preprocessing:
  • Normalization (mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])

Limitations

Please note that this model should be used as a supportive tool and not as a sole decision-maker for disease diagnosis.

Model Author

[Your Name/Organization]

Citation

If you use this model, please cite:

@software{vgg_cnn_rice_disease,
  title={VGG16-CNN Rice Disease Classification Model},
  version={0.1.0},
  year={2024}
}