Agrosight YOLOv11 Crop Disease

Model Overview

This is a fine-tuned YOLOv11 object detection model developed for Agrosight AI, an agricultural diagnostics platform designed for African smallholder farmers. It detects crop diseases, pests, and nutrient deficiencies across four major staple and cash crops.

This model serves as the core vision engine for the Agrosight AI platform, routing structured detection data to a localized language model for climate-smart advisory.

Dataset Details

The model was trained on a highly curated and localized agricultural dataset to ensure high performance in real-world African farming conditions.

• Total Images: 19,988 manually annotated images.
• Classes: 28 distinct classes covering diseases, pests, and healthy states.
• Supported Crops: Maize, Tomato, Rice, and Tea.

Performance Metrics

The model achieved the following validation metrics on the holdout set:

Usage Example

To deploy this model for inference using the ultralytics PyTorch package:

from ultralytics import YOLO

# Load the fine-tuned Agrosight YOLOv11 model
model = YOLO('hf://Nick-Maximillien/Agrosight-YOLOv11-Crop-Disease/best.pt')

# Perform inference on a sick crop image
results = model('sick_maize_leaf.jpg')

# Display results
for result in results:
    boxes = result.boxes
    for box in boxes:
        class_id = int(box.cls[0])
        confidence = float(box.conf[0])
        print(f"Detected: {model.names[class_id]} with {confidence:.2f} confidence")