Human Activity Recognition using Wi-Fi CSI

---

Domain: Human Activity Recognition, Wi-Fi Sensing

Tags: Wireless Signal, Machine Learning

Tools: Python

Status: Completed in May, 2022

Overview

This Human Activity Recognition using Wi-Fi CSI project classifies human activities by analyzing Wi-Fi Channel State Information (CSI), eliminating the need for imaging devices. Designed for semi-controlled environments, it leverages Wi-Fi signals to interpret activities like walking, sitting, and standing.

Additional Contributions

• Custom Dataset: Created a dataset with over 15 subjects and 800+ samples using microcontroller devices, contributing valuable data to the research community.
• Non-Intrusive Approach: Provides privacy-preserving activity recognition, beneficial for smart home environments and healthcare monitoring.

Practical Applications

The system supports non-intrusive activity recognition in semi-controlled environments, applicable in home automation, healthcare, and elderly care.