A Review - Smoke-Fire Detection and YOLO (You Only Look Once)

Abstract

Yolo’s deep learning algorithms make it possible to accurately detect smoke and fire in real time, making it a crucial tool for early fire detection and suppression. This Paper is a survey of the literature on fire detection over the previous three years (2020–2023) using the YOLO (you only look once) algorithm with Attention Mechanism . Due to its widespread use, YOLO has been the primary method of detection in the majority of published works. We have given a thorough review of the application of YOLO in smoke and fire detection in this research by comparing the published works using data sets, methodology, strategies and evaluating performance. To raise the detection rate and decrease the rate of false positives, the majority of works have trained or employed augmentation approaches and an attention model with various Image processing techniques.