Faculty

Ke Xu
Associate Professor

Email:l13025816@sjtu.edu.cn

Institute:Institute of Artificial Intelligence and Cognitive Security

Brief Introduction

Ke Xu, graduated with a PhD from Shanghai Jiao Tong University in 2019 and is currently serving as an Associate Researcher and Doctoral Supervisor in the School of Computer Science (School of Cyberspace Security) at Shanghai Jiao Tong University. The research focuses on digital media content security, mainly through basic theoretical research and technical implementation of video content security, video adversarial sample attacks and defenses, and related technologies. Published/accepted over 50 academic papers in domestic and international academic journals and conferences, including international academic journals such as TIFS, TMM, TCSVT, as well as international academic conferences such as ICASSP, ICME, and UAI. Authorized 14 invention patents and registered 3 software copyrights. He is a senior member of the Chinese Society of Image and Graphics, a member of the Digital Media Evidence and Security Special Committee, an IEEE member, and a TC member of international conferences such as APSIPA and ICSPS since 2020. In 2022, he served as the chairman of the ICSPS branch. Participated in the writing of the book "Principles and Application Practice of Digital Media Forensic Science", taught undergraduate courses such as "Computer Organization and System Architecture", "Computer Communication Network", and graduate courses such as "Multimedia Content Security".


Teaching Assignment

Taught undergraduate courses such as "Computer Organization and System Architecture", "Computer Communication Network", and graduate courses such as "Multimedia Content Security", and participated in the writing of the book "Principles and Application Practice of Digital Media Forensic Science".

Publications

Research has been conducted on multimedia content security, with the main research directions as follows: 
(1) Video anomaly event detection 
(2) Gait recognition and anonymization 
(3) Model adversarial attacks and detection

1.Yuting Mou, Ke Xu*, Xinghao Jiang, Tanfeng Sun:MV-guided deformable convolution network for compressed video action recognition with P-frames. Neurocomputing 649: 130770 (2025)

2.Chenting Xu; Ke Xu*; Xinghao Jiang; Tanfeng Sun. PLOVAD: Prompting Vision-Language Models for Open Vocabulary Video Anomaly Detection. TCSVT, 2025 (DOI: 10.1109/TCSVT.2025.3528108)

3.Ke Xu, Xinhao Jiang, Tanfeng Sun, Gait Recognition based on Local Graphical Skeleton Descriptor with Pairwise Similarity Network[J], IEEE Transactions on Multimedia, 2021, Vol.24, Page 3265-3275. DOI: 10.1109/TMM.2021.309580

4.Xu K, Jiang X, Sun T. Video Anomaly Detection and Localization Based on an Adaptive Intra-frame Classification Network[J],IEEE Transactions on Multimedia, 2020, 22(2):Page 394-406, DOI: 10.1109/TMM.2019.2929931

5.Xu K, Jiang X.*, Sun T. Anomaly Detection Based on Stacked Sparse Coding with Intraframe Classification Strategy[J]. IEEE Transactions on Multimedia, 2018, 20(5): 1062-1074.





Project Fund

1) National Natural Science Foundation of China General Fund,  (62372295) 
2) National Natural Science Foundation of China Youth Fund,  (62002220) 
3) 67th batch of general funding from China Postdoctoral Fund,  (2020M671124) 
4) 13th Special Support from China Postdoctoral Fund (Station),  (2020T130413)


Awards

AISafety-CVPR2022 Challenge Open Set Defense Track Runner up
CSIG FAT-AI 2022 Open Scene Face Combat Disguise Challenge Third Prize



Academic Service

Senior member of the Chinese Society of Image and Graphics, member of the Digital Media Evidence and Security Committee, IEEE member, and CCF member. From 2020 to present, served as a TC member of international conferences such as APSIPA and ICSPS. In 2022, served as the chairman of the ICSPS branch. Reviewer for journals such as TIFS, TMM, and conferences such as ICML, ACM MM.