Objective:

The goal of this project is to develop a machine learning model capable of recognizing motion sources (e.g., human movement, pets, or inanimate objects) based on WiFi signal data. The project will utilize a benchmark database for training and testing the model.

Background:

WiFi signals are increasingly used for indoor localization and motion detection due to their ubiquity and low cost. When a person or object moves within the range of a WiFi network, the signal's properties (like amplitude, phase, and Doppler shift) are altered. By analyzing these changes, it's possible to infer the type of motion occurring in the environment.

This project will leverage a pre-existing benchmark dataset, which contains WiFi signal data captured under various conditions. The dataset will be used to train a machine learning model to recognize different motion sources.