Patient care in long term home environments is one of the many applications that is currently being researched in the field of non-contact monitoring with radar. Most of the research has been done with data collected in very controlled situations. While this type of initial research is critical in the path to developing practical systems, there is a need to develop systems which deal with subjects behaving naturally, as they would in uncontrolled environments. Posture detection has not been performed in a robust manner using a single noncontact radar sensor, nor has it been implemented into a system that also performs activity classification. Identification of a subject’s posture is necessary for e.g. fall prevention; knowing that a subject who has a history of falling is standing may indicate that they are at a high risk of experiencing another fall. Hence in this thesis, we want to develop algorithms for radar-based Human Activity and Posture Classification. This fits in the broader context of all the mm-wave radar applications developed at IMEC for smart home/building/city, automotive and e-health.
The aim of this thesis is:
- To develop machine learning algorithms for Human Activity and Posture Classification using an existing radar prototype. The algorithms should be robust to noise as well as limited training data
- To develop algorithms for Radar data augmentation using generative adversarial networks
This work will be performed within a team working on all aspects of the radar, from RF to signal processing and machine learning. The work will include literature study, algorithm development and experimentation with a mm-wave radar.
The successful candidate(s) must show a good understanding and deep interest in signal processing, statistical learning and computer science. Proficiency with Matlab, Python and C/C++ is a must. Some knowledge of radar concepts is a plus.
• B. Erol, S. Z. Gurbuz and M. G. Amin, "GAN-based Synthetic Radar Micro-Doppler Augmentations for Improved Human Activity Recognition," 2019 IEEE Radar Conference (RadarConf), Boston, MA, USA, 2019, pp. 1-5. doi: 10.1109/RADAR.2019.8835589
• Y. Lin, J. Le Kernec, S. Yang, F. Fioranelli, O. Romain and Z. Zhao, "Human Activity Classification With Radar: Optimization and Noise Robustness With Iterative Convolutional Neural Networks Followed With Random Forests," in IEEE Sensors Journal, vol. 18, no. 23, pp. 9669-9681, 1 Dec.1, 2018.
• F. Luo, S. Poslad and E. Bodanese, "Kitchen Activity Detection for Healthcare using a Low-Power Radar-Enabled Sensor Network," ICC 2019 - 2019 IEEE International Conference on Communications (ICC), Shanghai, China, 2019, pp. 1-7.
Type of project: Combination of internship and thesis
Duration: 6 months
Preceded by optional summer internship (1 to 3 months) - the summer internship alone is not possible)
Supervising scientist: For further information or for application, contact Ali Gorji Daronkolaei (Ali.GorjiDaronkolaei@imec.be).