Millimeter wave radar data of people walking

Area of Science:

• Radar Signal Processing
• Human Gait Analysis
• Machine Learning

Background:

• Millimeter-wave (mmWave) Frequency Modulated Continuous Wave (FMCW) radar systems generate complex signals.
• Understanding human gait patterns is crucial for various applications, including healthcare and security.
• Evaluating radar data processing algorithms requires diverse and realistic datasets.

Purpose of the Study:

• To present a comprehensive dataset of mmWave FMCW radar signals during human walking activities.
• To facilitate research in human gait recognition using radar data.
• To provide a benchmark for evaluating the performance of different radar signal processing algorithms.

Main Methods:

• Acquisition of radar signals using a mmWave FMCW system in an indoor environment.
• Recording data from 29 subjects performing six distinct walking activities multiple times.
• Total of 231 acquisitions, with subjects walking naturally without specific constraints.

Main Results:

• A rich dataset containing complex radar signals corresponding to varied human gaits.
• The dataset includes multiple experiments per activity, ensuring data diversity.
• Signals capture natural walking patterns, not restricted by predefined movements.

Conclusions:

• The dataset is valuable for advancing human gait recognition research.
• It serves as an excellent resource for testing and comparing radar data processing techniques.
• This data supports the development of more robust and accurate radar-based human activity recognition systems.