YOLOv10-Fire and Smoke Detection

This project focuses on real-time detection and tracking of fires and smoke in video feeds using the latest YOLOv10 model. in this repository includes all the necessary code and resources to implement and demonstrate a robust fire and smoke detection system, showcasing the capabilities of the YOLOv10 model.

demo.mov

Hugging Face Demo

To see the model in action, you can test it through Hugging Face demo.

Data

The dataset used for this project consists of images containing fire and smoke,primarily collected from Roboflow. To ensure the quality and reliability of the dataset, we performed an extensive data cleaning process using the cleanvision library.

Data Cleaning Process

The data cleaning process was thoroughly documented in a Jupyter notebook. You can find the notebook, which includes all the code and results, here.

Data Download

You can download the dataset used in this project from the following link: Download Dataset.

cleanvision_report

Dataset Statistics

Total Number of Images After cleaning: 123,015
Distribution of Labels:

Training and Validation Metrics

Precision, Recall, and mAP50

To further evaluate the model's performance, we measure precision, recall, and mean Average Precision at 50% IoU (mAP50). These metrics help assess the accuracy and reliability of the model in detecting and classifying fire and smoke in images.

• Precision: The ratio of true positive detections to the total number of positive detections (both true and false). It indicates how many of the predicted fires or smoke instances are correct.
• Recall: The ratio of true positive detections to the total number of actual positive instances. It shows how well the model captures all the instances of fire and smoke.
• mAP50: The mean Average Precision at 50% Intersection over Union (IoU) is a comprehensive metric that combines precision and recall, providing a single score to evaluate the model's overall performance.

Model

The trained YOLOv10 model for fire and smoke detection is provided in this repository. You can download and use the pre-trained model to perform fire and smoke detection on your own images or videos.

The model weights can be downloaded from the link below: Download YOLOv10 Model Weights