柯剑鹏，男，1996年6月生，讲师，工学博士，硕士生导师。2018年6月毕业于西安电子科技大学，获英国威廉希尔公司学士学位；2020年6月毕业于武汉大学国家网络安全学院，获英国威廉希尔公司硕士学位；2025年6月毕业于武汉大学国家网络安全学院，获网络空间安全博士学位。
邮箱：jpke@wtu.edu.cn
1. 主要研究方向
人工智能安全，内容安全
2. 教学、科研项目
[1] 2024-2027，《面向多模态深度伪造生成的关键技术研究》（国家自然科学基金面上项目，参与）
[2] 2018-2020，《XXX信息产生、传播、影响与干预机理研究》（国家重点研发计划，参与）
3. 论文与专利
论文：
[1] Ke J, Wang L. DF-UDetector: An effective method towards robust deepfake detection via feature restoration[J]. Neural Networks, 2023, 160: 216-226.
[2] Ke J, Wang L. Understanding and leveraging vocoder fingerprints for synthetic speech attribution: J. Ke and L. Wang[J]. Applied Intelligence, 2025, 55(7): 673.
[3] Jianpeng K, Wenqi W, Kang Y, et al. Omni-detection of adversarial examples with diverse magnitudes[J]. China Communications, 2024, 21(12): 139-151.
[4] Ke J, Wang L, Ye A, et al. Combating multi-level adversarial text with pruning based adversarial training[C]//2022 International Joint Conference on Neural Networks (IJCNN). IEEE, 2022: 1-8.
[5] Ke J, Wang L, Liu J, et al. SFIA: Toward a Generalized Semantic‐Agnostic Method for Fake Image Attribution[J]. International Journal of Intelligent Systems, 2024, 2024(1): 7950247.
[6] Fu J, Wang L, Ke J, et al. GANAD: A GAN-based method for network anomaly detection[J]. World Wide Web, 2023, 26(5): 2727-2748.
[7] Liu J, Wang L, Wang R, et al. Exposing the forgery clues of DeepFakes via exploring the inconsistent expression cues[J]. International Journal of Intelligent Systems, 2025, 2025(1): 7945646.
[8] Liu J, Zhang M, Ke J, et al. Advshadow: Evading deepfake detection via adversarial shadow attack[C]//ICASSP 2024-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2024: 4640-4644.
[9] Fu J, Wang L, Ke J, et al. TSIDS: Spatial–temporal fusion gating multilayer perceptron for network intrusion detection[J]. Expert Systems with Applications, 2025, 263: 125687.
[10] Zeng B, Lai K, Ke J, et al. PatchFinger: A model fingerprinting scheme based on adversarial patch[C]//International Conference on Neural Information Processing. Singapore: Springer Nature Singapore, 2023: 68-80.
[11] 王丽娜, 柯剑鹏, 叶傲霜, 等. 基于多元特征的微博被劫持账户检测[J]. 武汉大学学报 (理学版), 2020, 66(2): 95-102.
专利：
[1] 一种基于三元特征的深度注意力谣言鉴别方法及装置，ZL201911074952.7
[2] 一种基于潜空间对抗净化的高鲁棒深度伪造检测方法及装置, CN202510467745.7
[3] 一种基于微生物遗传算法的黑盒对抗样本生成方法，ZL202010585924.8
4. 教科研奖励
5. 个人荣誉和指导学生获奖