UM-co-developed autonomous driving technology wins prize at international competition

A technology developed cooperatively between the University of Macau (UM) and the Shenzhen Institutes of Advanced Technology of the Chinese Academy of Sciences (SIAT) has been awarded second prize in an international algorithm competition titled “CVPR Security AI Challenger.”
The UM-SIAT project was part of a competition of 1,681 projects from teams around the world, with the UM-SIAT team the only one from Macau to win an award at the competition, which was jointly organized by the University of Illinois, Tsinghua University, and Ali Security.
According to a statement from the university, the UM-SIAT project tested a new algorithm that can achieve better results than any other model robustness evaluation method that has emerged in recent years.
While the model of deep learning has been integrated into the fields of perception, decision-making, and the control of autonomous driving, previously presented deep learning models have presented severe problems concerning robustness. Among these is the vulnerability of self-driving visual perception systems to malicious attacks on adversarial samples, additions to image data, and subtle disturbances that are difficult for humans to recognize through their senses, all of which can cause models to make incorrect judgments.
The new model from UM-SIAT establishes an effective model robustness evaluation mechanism, ensuring a safe and viable autonomous driving system that fixes many of the misperceptions occurring in other models that pose a significant safety risk to operational self-driving vehicles.
Accompanying the competition entry, the UM-SIAT has presented a paper further explaining the new model to the international conference dedicated to the field of artificial intelligence.
The academic paper is authored by a student, Yu Yunrui, under the guidance of Professor Xu Cheng-Zhong, dean of the Faculty of Science and Technology of the UM and the academic leader of the State Key Laboratory of Smart City Internet of Things. The paper is co-authored by Dr. Gao Xitong, an assistant researcher of SIAT, also under the guidance of Professor Xu.