Wi-Chat represents the first system to combine Large Language Models with Wi-Fi signals to recognize human activities, creating new possibilities for security monitoring.

• Processes raw Wi-Fi signals to detect and identify human movements
• Integrates physical model knowledge with LLM capabilities for real-world signal interpretation
• Enables non-invasive monitoring without cameras or wearable devices
• Creates new opportunities for security surveillance and intrusion detection

This breakthrough demonstrates how LLMs can extend beyond text to interpret physical signals, offering security professionals powerful new tools for environment monitoring without privacy-invasive cameras.

Wi-Chat: Large Language Model Powered Wi-Fi Sensing