Adaptive Scheme for Crowd Counting Using off-the-Shelf Wireless Routers

摘要

Since the outbreak of the world-wide novel coronavirus pandemic,crowd counting in public areas, such as in shopping centers and in commercialstreets, has gained popularity among public health administrations for preventingthe crowds from gathering. In this paper, we propose a novel adaptive method forcrowd counting based on Wi-Fi channel state information (CSI) by using commoncommercial wireless routers. Compared with previous researches on device-freecrowd counting, our proposed method is more adaptive to the change of environmentand can achieve high accuracy of crowd count estimation. Because the distancebetween access point (AP) and monitor point (MP) is typically non-fixed inreal-world applications, the strength of received signals varies and makes the traditionalamplitude-related models to perform poorly in different environments. Inorder to achieve adaptivity of the crowd count estimation model, we used convolutionalneural network (ConvNet) to extract features from correlation coefficientmatrix of subcarriers which are insensitive to the change of received signal strength.We conducted experiments in university classroom settings and our model achievedan overall accuracy of 97.79% in estimating a variable number of participants.