Real-time 3D Human Motion Detection and Visualization using WiFi CSI

WiFi-3D-Fusion is an open-source project that leverages Channel State Information (CSI) from local Wi-Fi to perform real-time human motion detection and 3D visualization. Supporting both ESP32-CSI and Nexmon data acquisition, it employs advanced CNNs for person detection and tracking, including multi-person identification and re-identification. A continuous learning pipeline allows the model to automatically improve during operation. Visualization is offered through both a web interface and a terminal-based pipeline. Optional integrations with Person-in-WiFi-3D, NeRF², and 3D Wi-Fi Scanner are also provided.