Skip to content Skip to navigation

Academic Events

[Invited Talk] Professor Yuhui Zheng from Nanjing University of Information Science and Technology

Professor Yuhui Zheng, Nanjing University of Information Science and Technology, gave a talk entitled " Cross-platform video surveillance pedestrian re-identification for social public safety " at the School of Computer and Electronic Information on Nov. 25, 2021. The academic presentation was hosted by Professor Junsheng Zhou, Associate Dean of the School of Computer and Electronic Information.

Yuhui Zheng is a professor at Nanjing University of Information Science and Technology, winner of the Jiangsu Outstanding Youth Fund, a Minjiang Scholar, a young and middle-aged academic leader of the "Qinglan Project" in Jiangsu Province, member of the Chinese Society of Cognitive Science-Cognitive and Brain-inspired Computing Committee, and standing committee member of the Jiangsu Artificial Intelligence Society-Intelligent Perception Technology and Application Committee. He has presided over 4 national-level projects, including key projects of the National Natural Science Foundation of China, and international (regional) cooperative research and exchange projects. Now he is mainly engaged in computer vision perception and understanding methods and their application in security.

. To achieve an in-depth understanding of social public safety events, Professor Zheng's team has researched multi-source video data event correlation and cross-platform data content online correlation in recent years around the problem of cross-platform video data association with pedestrian re-identification technology. Among them, a cross-resolution pedestrian re-recognition method based on deep and high-resolution representation learning is proposed, which overcomes the difficulty of re-recognition caused by the different resolutions of multi-source video images. A brightness-adapted pedestrian re-recognition method is proposed, which effectively improves the accuracy of cross-platform video image re-recognition.