Project Report on finger gesture recognition using WiFi

This project is based on fine-grained finger gesture recognition by using a single WiFi device without requiring user to wear any sensors. The system senses and identifies subtle movements of finger gestures by examining the unique patterns exhibited in the detailed CSI. The core of the system consists of the Environmental Noise Removal and the Gesture Pattern Extraction. Environmental Noise Removal utilizes Wavelet Based Denoising to remove the noise by decomposing signals into approximation coefficients and detail coefficients. Next, the system performs gesture pattern extraction by utilizing Principal Component Identification. At last, the system recognizes gestures by calculating the similarity between the extracted CSI pattern and the pre- constructed gesture profiles. Our project identifies gesture using two methods: PCA and Eigen vectors. It then compares the performance using both the methods. I will give more details on both methods when we talk. P.S. - The project DOES NOT use multi-path mitigation and critical subcarrier selection mentioned in the architectural diagram.