LAUSR

School violent behaviors is a serious problem that greatly affects the healthy growth of the youth and the children. Current prevention measures mainly depend on propaganda and self-report. So far, there is still no effective solution that can automatically detect the violent behaviors. In this paper, we take the first attempt to build a ubiquitous passive violence detection system, WiVi, based on the commercial WiFi infrastructure. To capture the patterns of the complicated violent behaviors actions, besides the time-series features used in existing activity recognition works, WiVi also leverages the correlated features extracted from combined subcarriers, to take fully advantages of Channel State Information. WiVi integrates a feature fusion method to select the appropriate features for the classification model in different scenarios. We implement and evaluate WiVi in various real-world environments. The experimental results show that the recall and specificity that WiVi can achieve 93.46% and 93.57% respectively.