Subject

The advent of low-cost and small form factor multi-input multi-output (MIMO) Radar sensors has enabled numerous new applications that are well beyond those that are traditionally associated with Radar. One such application, motivated by the unique privacy-preserving advantage that Radar offers over traditional cameras, is the identification of individual people by their radar reflection. The goal in this master thesis project is to explore and possibly further the limits of the capabilities of current commercial of the shelf (COTS) Radar sensors for the application of gait-based person identification. A first investigation into this topic, from which the student can benefit, has already happened at imec and has shown that using micro-Doppler imaging and temporally processing neural networks is a viable approach. However, it remains an open question on how well such a system could perform under varying subject and environmental circumstances in real-life, as well as how to safeguard against such variances.

Kind of work

The work includes:
• Setting up a radar capturing system, built from COTS components,
• Measurement and analysis of an extensive dataset,
• Exploration of capture segmentation and preprocessing steps,
• Proposing a neural network architecture for real-time classification of the preprocessed images,
• Validating the proposed architecture.

Framework of the Thesis

Type:
- Master Thesis internship (6 months)
- Preceded by optional summer internship (max 3 months)

Responsible scientist(s): Lars Keuninckx (lars.keuninckx@imec.be)

Expected Student Profile

The successful candidate must be competent at Python coding and have some experience with relevant machine learning and AI toolkits (such as Scipy, Statsmodels, Sci-Kit-learn, PyTorch, Keras etc.). A good understanding of signal processing is also required. Prior knowledge of radar concepts is a plus.

Interested students can already get a feel for this subject from these papers:

1. B. Vandersmissen et al., "Indoor Person Identification Using a Low-Power FMCW Radar", in IEEE Transactions on Geoscience and Remote Sensing, vol. 56, no. 7, pp. 3941-3952, July 2018, doi: 10.1109/TGRS.2018.2816812.
2. A. Jalalvand, B. Vandersmissen, W. De Neve and E. Mannens, "Radar Signal Processing for Human Identification by Means of Reservoir Computing Networks" 2019 IEEE Radar Conference (RadarConf), 2019, pp. 1-6, doi: 10.1109/RADAR.2019.8835753.