Students at XJTLU-JITRI Academy of Industrial Technology recently won four prizes in the second JITRI Innovation Cup competition. The students won one of the three first prizes, one of the five second prizes and two of the twelve third prizes.
The competition was hosted by the Yangtze River Delta National Technology Innovation Center, and this year received a total of 20 teams with 108 participants. The committee selected 20 research topics that address major industrial needs in the Yangtze River Delta region, and invited domestic universities to assemble interdisciplinary research teams to tackle them.
Yi Zhou
Multipath and ghost images
The first prize was awarded to a team of six postgraduate students from the XJTLU-JITRI Academy of Industrial Technology for their method to improve deep radar perception and mitigate multipath effects for autonomous vehicles.
Yi Zhou, the team leader, says: “Multipath is a phenomenon where a wave from a target travels to a detector through two or more paths. Because of multipath propagation, the radar receives both direct reflections and indirect time-shifted reflections.
“Sometimes the multipath effects are materialised in fake impressions of the true target, known as ghost images. For example, the radar might mistake a bicycle for a pedestrian, which may cause safety issues.”
To solve the problem, their team proposed a probabilistic detection method, and refined their research scenario.
“Our research mainly focuses on two kinds of moving targets – pedestrians and bicycles, especially the ghost images reflected on walls when they are passing by in the city,” says Zhou. “With the help of our mentors, we found a research direction which not only has academic value but also meets the actual industrial needs.”
According to the team, in recent years, the international first-tier automobile companies have obtained several patents in ghost image detection, but there is little research in China. If there’s a breakthrough in this field, it can greatly contribute to the development of intelligent transportation and vehicles in China.
Academic-industry collaboration
The project, which is still ongoing, is supervised by Dr Limin Yu, Associate Professor at XJTLU’s Department of Electrical and Electronic Engineering, and Dr Yutao Yue, industrial mentor at JITRI’s Institute of Deep Perception Technology.
Ningfei Song, one of the team members, says Dr Yue provided them with the industry perspective on whether their research has practical relevance. “With equipment provided by the industry, our ideas no longer remain on paper, but can be tested and verified in real projects in a timely manner.”
Zhou adds: “Dr Yu inspired us from the academic perspective, and the experienced engineers at the institute gave us technical guidance.”
The JITRI Innovation Cup aims to strengthen the connection between industrial demands and scientific and educational resources in the Yangtze River Delta region. Dr Yu says: “The competition provides a platform and enhances students’ research skills. It focuses on scientific and industrial needs, and provides a model of academic-industry collaboration.”
In the future, Zhou and other students at the XJTLU-JITRI Academy of Industrial Technology will continue to work on this project, and apply algorithms to improve radar target identification and ghost image reduction.
The paper, ‘Towards deep radar perception for autonomous driving: datasets, methods, and challenges’ was published in the journal Sensors and can be found here. The team also created a homepage for researchers to browse.
By Yuxiao Zhang
Translated by Ziying Shi
Edited by Xinmin Han and Patricia Pieterse
Photos by XJTLU-JITRI Academy of Industrial Technology