Description of field of research:

mmWave radio sensing is a new technique that uses the radio signals from radars to study environment variations [1]. Any object (e.g., fingers) moving in the wireless environment will cause variations (e.g., amplitude, phase and doppler) in the received radio signal. This project aims to realise a networked embedded mmWave radar-based gesture recognition system via deep learning for device-free ubiquitous hand gesture monitoring. Pilot studies show different handrub gestures can be sensed and recognised by analysing the radio signal variations in the receiver [2]. This project will carry out further research with on the signal modelling and simulation, novel deep learning model design, prototyping with commercial off-the-shelf mmWave radars and on-field experimental validations.

Research Area

Networked embedded sensing | Radio sensing | mmWave radar | Edge AI

You will work with a team of Postdoctoral fellows and PhD students to develop novel deep learning models for mmWave radar-based gesture recognition, prototype the proposed models with commercial off-the-shelf mmWave radars and conduct on-field experimental model validations.

A robust, accurate and user-friendly hand gesture recognition systems based on mmWave radars.

[1] Nirmal I; Khamis A; Hassan M; Hu W; Zhu X, 2021, 'Deep Learning for Radio-Based Human Sensing: Recent Advances and Future Directions', IEEE Communications Surveys and Tutorials, vol. 23, pp. 995 - 1019

[2] Abdelwahed Khamis, Branislav Kusy, Chun Tung Chou, Mary-Louise McLaws, and Wen Hu. 2020. RFWash: a weakly supervised tracking of hand hygiene technique. Proceedings of the 18th Conference on Embedded Networked Sensor Systems (SenSys). Pp. 572–584.