Robust Recognition of Indoor Educational Activity Based on Wi-Fi Signals with SSGAN

Robust Recognition of Indoor Educational Activity Based on Wi-Fi Signals with SSGAN

Wi-Fi-based activity recognition has been widely used in many fields. However, there is no research dedicated to educational activity recognition, and a lot of works using Wi-Fi signals for human activity recognition does not have generalization performance for left-out users whose CSI data are not training in model. In order to solve the above problems, we propose an educational activity recognition model which can effectively target at left-out user. In addition, we considered the generalization performance of the model when facing the activities with different directions and different scenes. The experimental results show that our model has excellent performance in the above situations, and also has higher recognition accuracy in the face of left-out user.