Yolov10 Fire And Smoke Detection

🚀 YOLOv10: Real-Time Fire and Smoke Detection

This repository offers a YOLOv10 model trained for real - time fire and smoke detection. It leverages the Ultralytics YOLO framework to achieve high - accuracy and efficient object detection. Users can fine - tune confidence and IoU thresholds for optimal results.

🚀 Quick Start

This model is hosted on Hugging Face and can be accessed via the Inference Widget or programmatically using the Hugging Face Transformers pipeline.

✨ Features

• The model uses the Ultralytics YOLO framework for high - accuracy and efficient object detection.
• Users can adjust the confidence and IoU thresholds for optimal detection results.

📦 Installation

No specific installation steps are provided in the original document, so this section is skipped.

💻 Usage Examples

Basic Usage

Inference Widget

Upload an image to the widget below and adjust the following:

• Confidence Threshold: Minimum confidence level for predictions (default: 0.25).
• IoU Threshold: Minimum IoU level for object matching (default: 0.45).
• Image Size: Resize input image (default: 640x640).

Usage with Python

To use the model programmatically:

import torch
from ultralytics import YOLO
from PIL import Image

# Load the model
model_path = "pytorch_model.bin"
state_dict = torch.load(model_path, map_location="cpu")

# Initialize the YOLO model
model = YOLO()  # Replace with the correct YOLO class
model.load_state_dict(state_dict)

# Run inference
image = Image.open("path/to/image.jpg")
results = model.predict(image, conf=0.25, iou=0.45)
results.show()

📚 Documentation

Model Details

📄 License

The model is released under the apache - 2.0 license.