360 degree perception of a vehicle surrounding using a network of radars

While multitarget tracking (MTT) is challenging in a single radar, high performance is achievable by a network of multiple-input-multiple-output (MIMO) radars. Such multi-radar systems have become pervasive in self-driving applications for object detection and tracking purposes.

IMEC has successfully developed multi-radar MTT algorithms based on the state-of-the-art Probability Hypothesis Density (PHD) filter [1] and the consensus-based track fusion [2]. The PHD filter is a state-of-the-art tracking algorithm based on the theory of random-finite sets (RFS).

The goal of this master thesis is to further develop the MTT and fusion algorithms based on the cardinalized PHD (CPHD) filter [3]. In parallel, a data-driven deep-learning tracking model is developed [4,5].  For the end-to-end approach, the idea is to get insights from state-of-the-art deep learning models for the camera- and lidar-based tracking and detection, and adapt those to fit in radar data-based detection and tracking tasks and compare these to the CPHD-based results. 

The work steps will include the following: 

• Implementing the CPHD filter for MIMO radar;
• Track fusion to obtain 360^o perception in the vehicle;
• Implementing an end-to-end deep learning approach for MIMOM radar detection and tracking.
  
  

Required qualifications:

• Following an MSc in a field related to one or more of the following: Electrical engineering, Computer Science, or Applied Computer Science.
• Experience with signal processing, and machine learning. Some knowledge of radar concepts is a plus.
• Strong programming skills (Python).
• Interest in developing state-of-the-art Machine Learning methods and conducting experiments.
• Ability to write scientific reports and communicate research results at conferences in English.

Type: 

- Master Thesis internship (6 months) 

- Preceded by optional summer internship (max 3 months) 

Responsible scientist(s): 

Hichem Sahli (hichem.sahli@imec.be); 

Seyed Hamed Javadi (hamed.javadi@imec.be)  

References:

1. B.-N. Vo and W.-K. Ma, “The Gaussian Mixture Probability Hypothesis Density Filter,” IEEE Transactions on Signal Processing, vol. 54, no. 11, pp. 4091–4104, 2006. 
2. G. Battistelli, L. Chisci, C. Fantacci, A. Farina and A. Graziano, "Consensus CPHD Filter for Distributed Multitarget Tracking," in IEEE Journal of Selected Topics in Signal Processing, vol. 7, no. 3, pp. 508-520, June 2013.
3. B.-T. Vo, B.-N. Vo and A. Cantoni, “Analytic implementations of the cardinalized probability hypothesis density filter,” IEEE Trans. on Signal Processing, vol. 55, no. 7, pp. 3553-3567, 2007.
4. A. Palffy, J. Dong, J. F. P. Kooij and D. M. Gavrila, "CNN Based Road User Detection Using the 3D Radar Cube," in IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 1263-1270, April 2020.