The elite International Radar Symposium took place in Germany in June this year which was attended by radar professionals from popular space and defence companies from all over the globe including School of Engineering students, Ana Stroescu, Dom Phippen and colleagues from Microwave Integrated Systems Laboratory (MISL), who presented their research papers.
Ana and her Supervisors Professor Marina Gashinova and Professor Mike Cherniakov from the University of Birmingham's MISL group, a world-class radar research group and one of the largest and best UK university groups in Europe, submitted a conference paper analysing the role of artificial intelligence in the use of high resolution radar imagery in autonomous driving systems. To their delight they won ‘Best Paper Award’ for the Classification of High Resolution Automotive Radar Imagery for Autonomous Driving Based on Deep Neural Networks and winning €2000 cash prize! To add to the success, Ana was also nominated in the top 5 for the ‘Young Scientist Award’!
Ana’s research relates to the classification of radar targets in high resolution, low-THz radar imagery for autonomous driving and using deep neural networks. This was her first published paper which was written last year, whilst on a Postgraduate Masters Research studies (MRes) course she completed prior to the application for doctoral studies.
The abstract of the paper explains Ana’s research further: "Recently, there has been an increase in research for deep neural networks that perform object classification for self-driving vehicles using electro-optical sensors. Public optical datasets and classification algorithms that enable such development already exist, however, only radars can provide robust sensing in adverse conditions, when the optical systems may fail. The development of high resolution radar is necessary in order to approach radar imagery classification with neural networks in a similar way with optical images”.
During the presentation, the team highlighted a method of classification of six different roadside targets in low-THz imaging radar, using convolutional neural networks. The present results confirmed that neural networks can also be successfully employed for low-THz radar imagery classification with high resolution and this method can be applicable for implementing an all-weather sensing system for autonomous vehicles.
Ana was funded by Jaguar Land Rover during her MRes studies and is currently employed at the University of Birmingham. Her research focuses on computer vision and machine learning applied to incredibly high resolution and high frequency radar data. This particular radar has been showcased in a research video.
The paper presented by Ana at the conference focused on using neural networks on this radar data to teach a computer to recognise different kinds of road objects and users. This research is particularly important for identifying road users for future autonomous driving, which ensures the safety of different vulnerable road users especially in difficult lighting and/or extreme weather conditions where cameras or lidars may struggle.
To add to the outstanding achievement, Hensoldt uploaded a Linkedin post to highlight Ana’s research success which created further engagement.