
职称:副教授
个人简介:
刘心晶,博士,副教授。联系方式:liuxinjing@nwu.edu.cn
2026年6月于西安电子科技大学获得网络空间安全博士学位,博士期间担任香港城市大学计算机科学系科研助理1年,2026年9月起在西北大学计算机学院任教。
近年来,围绕人工智能模型的安全威胁与防护机制开展了系统性研究,相关成果发表在IEEE Transactions on Information Forensics and Security(TIFS)、IEEE Transactions on Dependable and Secure Computing(TDSC)、AAAI、ESORICS、中国科学:信息科学、IEEE Internet of Things Journal等国内外重要学术期刊及会议,以第一/通讯作者发表学术论文10篇,其中CCF推荐A类论文6篇;获授权国家发明专利2项、软件著作权1项;作为核心成员参与国家自然科学基金、国家重点研发计划等科研项目5项。
研究方向:人工智能安全;模型所有权保护;对抗攻防
论文成果:
[1] Xinjing Liu, Yilong Yang*, Taifeng Liu, et al. PROTheft: A Projector-Based Model Extraction Attack in The Physical World. IEEE Transactions on Information Forensics and Security, 2026. [TIFS, CCF-A]
[2] Xinjing Liu, Taifeng Liu, Hao Yang, et al. AttMark: Attention Based Watermarking for Neural Networks. IEEE Transactions on Dependable and Secure Computing, 2024. [TDSC, CCF-A]
[3] Xinjing Liu, Zhuo Ma, Yang Liu, et al. SeInspect: Defending Model Stealing via Heterogeneous Semantic Inspection. European Symposium on Research in Computer Security, 2022: 610–630. [ESORICS, CCF-B]
[4] Zhuo Ma, Xinjing Liu, Yang Liu, et al. DivTheft: An Ensemble Model Stealing Attack by Divide-and-Conquer. IEEE Transactions on Dependable and Secure Computing, 2023: 1–13. [TDSC, CCF-A]
[5] 刘心晶, 王珠珠*, 孟宪佳*, 等. 类增量学习中基于语义强化的可持续对抗攻击. 中国科学:信息科学, 2026.
[6] Taifeng Liu, Xinjing Liu*, Liangqiu Dong, et al. Improving Sustainability of Adversarial Examples in Class-Incremental Learning. Proceedings of the AAAI Conference on Artificial Intelligence, 2026. [AAAI, CCF-A]
[7] Yilong Yang, Xinjing Liu*, Zefeng Wu, et al. Catch Me If You Can: Retain High Stealthiness and Durability of Backdoor Attack in Federated Learning. IEEE Transactions on Information Forensics and Security, 2025. [TIFS, CCF-A]
[8] Xinjing Liu, Taifeng Liu*, Hao Yang, et al. Model Stealing Detection for IoT Services Based on Multi-Dimensional Features. IEEE Internet of Things Journal, 2024, 11(24): 39183–39194. [IoT-J]
[9] Teng Fei, Xinjing Liu, Yanjun Zhang, et al. MarkErase: Defeating Entangled Watermarks in Model Extraction Attacks. Advances in Knowledge Discovery and Data Mining, 2025: 78–90. [PAKDD, CCF-C]
[10] Taifeng Liu, Chao Yang, Xinjing Liu, Ruidong Han, et al. RPAU: Fooling the Eyes of UAVs via Physical Adversarial Patches. IEEE Transactions on Intelligent Transportation Systems, 2024, 25(3): 2586–2598. [TITS]
[11] Hao Yang, Zhuo Ma, Yang Liu, Xinjing Liu, et al. Updates Leakage Attack against Private Graph Split Learning. International Conference on Algorithms and Architectures for Parallel Processing, 2024. [ICA3PP, CCF-C]
[12] Zhuo Ma, Yilong Yang, Yang Liu, Xinjing Liu, et al. Mitigate Noisy Data for Smart IoT via GAN Based Machine Unlearning. Science China Information Sciences, 2023. [CCF-A]
[13] Zhuo Ma, Yilong Yang, Yang Liu, Tong Yang, Xinjing Liu, et al. Need for Speed: Taming Backdoor Attacks with Speed and Precision. IEEE Symposium on Security and Privacy, 2024: 1217–1235. [S&P, CCF-A]
[14] Taifeng Liu, Yang Liu, Zhuo Ma, Tong Yang, Xinjing Liu, et al. L-HAWK: A Controllable Physical Adversarial Patch Against a Long-Distance Target. Network and Distributed System Security Symposium, 2024. [NDSS, CCF-A]
[15] Zhuo Ma, Yilong Yang, Bin Xiao*, Yang Liu, Xinjing Liu, et al. Sniffer: A Novel Model Type Detection System against Machine-Learning-as-a-Service Platforms. Proceedings of the VLDB Endowment, 2024: 3942–3945. [VLDB, CCF-A]
[16] Zhuo Ma, Jiayu Jin, Yang Liu, Yilong Yang, Xinjing Liu, et al. SafeLead: Detecting and Excluding Random STS Attack in UWB Ranging System. IEEE International Conference on Computer Communications, 2025: 1–10. [INFOCOM, CCF-A]
获奖情况:
[1] 中国计算机网络与信息安全会议(CCNIS 2025)大会最佳论文、分论坛最佳论文、最佳表现奖。
[2] International Conference on Algorithms and Architectures for Parallel Processing(ICA3PP)最佳论文奖。
[3] 第四届“华为杯”中国研究生网络安全创新大赛二等奖。