Multi-Sensor Data Collection (Radar + Azure Kinect)

Collecting labelled multimodal human-sensing data is fiddly: multiple sensors, different SDKs, different clocks. This tool makes it repeatable — a modular GUI that synchronises TI mmWave radar (AWR2243) captures with Azure Kinect RGB-D recordings and structures the output for downstream labelling and training.

It underpins the multimodal datasets used in my PhD research on radar-based human activity recognition.

Details

  • Synchronised capture across TI mmWave Studio (driven via a Lua TCP bridge) and the Azure Kinect SDK
  • Structured, labelled recording sessions for repeatable experiments
  • Built with Python, PySimpleGUI, and OpenCV

Code: github.com/keniel123/multi-sensor-data-collection