/CSIKit-BMI

Primary LanguagePython

CSIKit-BMI

This is the official implementation for BMEye: Public Health-Oriented Body Mass Index Monitoring using Commodity WiFi accepted at WF-IoT '23.

Heatmaps in the dataset are 2-D images converted using code in Section CSI-Heatmap conversion.

[Paper]

System_overview

🔧 Requirements

Requirements are given in requirements.txt file

To install all packages and libraries needed, run pip install -r requirements.txt

⚡ CSI- Heatmap Conversion

⚡ Train given CSI Heatmaps:

  1. The dataset folder contains individual class Heatmaps in .zip files. Unzip them and place them there only.
  2. Run 'resnet50-csi-pytorch.py' to train with above data.

⚡ Train custom CSI Samples:

  1. To train on custom CSI data, convert CSI .pcap/.dat files using code in Section CSI-Heatmap conversion.