Automatic detection of indoor occupancy based on improved YOLOv5 model

Abstract

Indoor occupancy detection is essential for energy efficiency control and Coronavirus Disease 2019 traceability. The number and location of people can be accurately identified and determined through classroom surveillance video analysis. This information is used to manage environmental equipment such as HVAC and lighting systems to reduce energy use. However, the mainstream one-stage YOLO algorithm still uses an anchor-based mechanism and couples detection heads to predict. This results in slow model convergence and poor detection performance for densely occluded targets. Therefore, this paper proposed a novel decoupled anchor-free VariFocal loss convolutional network algorithm DFV-YOLOv5 for occupancy detection to tackle these problems. The proposed method uses the YOLOv5 algorithm as a baseline. It uses the anchor-free mechanism to reduce the number of design parameters needing heuristic tuning. Afterwards, to reduce the coupling of the model, speed up the model's convergence ability, and improve the model detection performance, the detection head is decoupled based on the YOLOv5 model. It can resolve the conflict between classification and regression tasks. In addition, we use the VariFocal loss to assign more weights to difficult data points to optimize the class imbalance problem and use the training target q to measure positive samples, treating positive and negative samples asymmetrically. The total loss function is redesigned, the L 1 loss is increased, and the ablation experiment verifies the effect of the improved loss. By applying a hybrid activation function of the sigmoid linear unit and rectified linear unit, we improved the model's nonlinear representation and reduced the model's inference time. Finally, a classroom dataset was constructed to validate the occupancy detection performance of the model. The proposed model was compared with mainstream target detection models regarding average mean precision, memory allocation, execution time, and the number of parameters on the VOC2012, CrowdHuman and self-built datasets. The experimental results show that the method significantly improves the detection accuracy and robustness, shortens the inference time, and proves the practicality of the algorithm in occupancy detection compared with the mainstream target detection model and related variants of the model.

Forest Fire Smoke Detection Based on Deep Learning Approaches and Unmanned Aerial Vehicle Images

“…Detection frameworks have been the subject of further research into developing specific methods for interior and exterior item detection and classification that may be applied to detect flames and smoke. One such study is [ 44 ], which presented a unique architecture for identifying items within occupied spaces. By utilizing the anchor-free technique for variable minimization and the VariFocal loss for data weighing, this suggested design made optimal use of the YOLOv5 model.…”

Using the improved YOLOv5 network to detect the real-time and effective position of milk and construct the milk data set

“…Zhong Yuan et al [17] added an attention mechanism based on YOLOv5 to highlight important information in the feature map, weaken secondary information, and enhance the recognition ability of the network. Wang Chao et al [18] proposed a convolution algorithm DFV-YOLOv5 based on decoupled anchor-free and zoom loss to speed up the convergence efficiency of the model, and used the zoom loss technology to improve the detection accuracy of occluded targets.…”

Hold surrounding’s key-you only look once version 7: a real-time pedestrian and vehicle detection algorithm in the low-signal-to-noise ratio infrared image

“…The core of environment perception based on infrared sensors is object detection 6 . Object detection algorithms are widely used in pedestrian and vehicle detection and are mainly divided into object detection algorithms using traditional machine learning (ML) and object detection algorithms using deep learning (DL) 7 .…”

“…The core of environment perception based on infrared sensors is object detection. 6 Object detection algorithms are widely used in pedestrian and vehicle detection and are mainly divided into object detection algorithms using traditional machine learning (ML) and object detection algorithms using deep learning (DL). 7 Dalal et al 8 used the histogram of oriented gradient (HOG) algorithm and the support vector machine (SVM) algorithm for pedestrian detection, which is widely used in autonomous driving tasks.…”