Fire and Smoke Detection Using Fine-Tuned YOLOv8 and YOLOv7 Deep Models

Fire and Smoke Detection Using Fine-Tuned YOLOv8 and YOLOv7 Deep Models

2024 | Mohamed Chetoui and Moulay A. Akhloufi
This paper presents a fine-tuned YOLOv8 model for fire and smoke detection in various locations, including forests, cities, and industrial zones. The authors utilized a dataset comprising over 11,000 images of smoke and fires to train and optimize the YOLOv8 model. The YOLOv8 model achieved high precision, recall, and mean average precision (mAP:50) scores of 92.6%, 83.7%, and 95.2%, respectively. The results were compared with other models such as YOLOv7, YOLOv6, Faster-RCNN, and DETR, showing superior performance in terms of precision, recall, and mAP:50. The proposed model was tested on different conditions, including day and night, and in various environments, demonstrating its robustness and adaptability. The study highlights the potential of deep learning models in enhancing fire and smoke detection, particularly in real-time applications, and suggests that the YOLOv8 model can be effectively deployed on edge devices for efficient and accurate fire safety monitoring.This paper presents a fine-tuned YOLOv8 model for fire and smoke detection in various locations, including forests, cities, and industrial zones. The authors utilized a dataset comprising over 11,000 images of smoke and fires to train and optimize the YOLOv8 model. The YOLOv8 model achieved high precision, recall, and mean average precision (mAP:50) scores of 92.6%, 83.7%, and 95.2%, respectively. The results were compared with other models such as YOLOv7, YOLOv6, Faster-RCNN, and DETR, showing superior performance in terms of precision, recall, and mAP:50. The proposed model was tested on different conditions, including day and night, and in various environments, demonstrating its robustness and adaptability. The study highlights the potential of deep learning models in enhancing fire and smoke detection, particularly in real-time applications, and suggests that the YOLOv8 model can be effectively deployed on edge devices for efficient and accurate fire safety monitoring.
Reach us at info@study.space
Understanding Fire and Smoke Detection Using Fine-Tuned YOLOv8 and YOLOv7 Deep Models