A lightweight fire detection model integrating attention mechanism
Based on visual information,fire detection is of great significance to fire protection work.However,most of the meth-ods proposed by relevant research institutions at this stage are based on high-performance hardware devices,which limits the practical deployment and application of relevant results.In re-sponse to this,this paper uses ShuffleNetv2 network as the main backbone to construct a lightweight model based on YOLOv5 tar-get detection model,and introduces the SIoU loss function to im-prove the positioning accuracy of the model's target box.Addition-ally,the Shuffle Attention module is added to the model to im-prove its recognition accuracy of flame targets in complex environ-ments.Experiments have shown that compared to the original YOLOv5 model,the improved model not only achieves better recognition results but also reduces the parameter count by 54.2%and improves detection speed by 40.5%.Finally,the model is de-ployed to embedded devices to verify its application efficiency,and the results show that while maintaining recognition perfor-mance,the model achieves a detection speed of 32 f/s.
NETL
NSTL
万方数据
