Reading notes
Summary of notes taken by us reading resources provided by the client (TI)
List of Documents / Websites:
Summary
1. Machine Learning on the Edge with the mmWave Radar Device IWRL6432
- To be complete
2. Human Posture Capturing with Millimeter Wave Radars
- To be complete
3. Surface Classification User Guide
- Overview: A guide focused on classifying surfaces using mmWave radar. It provides an out-of-the-box demo and steps to get started.
- Key Topics:
- Surface Classification Demo Video: Radar-based classification of different surfaces using either the IWR6843AOPEVM or IWRL6432BOOST EVM.
- Out-of-the-Box Demo: Instructions to run the demo provided by TI for surface classification.
- Developer’s Guide
- Actionable Stuffs:
- Follow the Out Of Box Demo User Guide to set up and run a surface classification test.
4. Human Activity Recognition Using Millimetre-Wave Radars With Machine Learning
- Overview: A thesis on human activity recognition using mmWave radar sensors. It discusses the use of machine learning algorithms to classify human postures and vital signs.
- Key Topics:
- Extracting Usable Data: They used two radar sensors to verify each other’s data to increase the accuracy of the model by reducing the input noise.
- Detecting human postures: They created a model which could estimate the position of a person’s head, shoulders, elbows, hips and knees.
- Detecting vital signs: They used a radar sensor to detect the heart rate of a person whilst they’re moving.
- Error Rate: They achieved very low error rates. It was only 5.4% for a person’s heartrate whilst excersing on a treadmill, for example.
- Privacy: The paper was very positive about the privacy benefits of using radar sensors over cameras - especially in health and social care.