主要研究方向：人工智能安全、多模态学习、智能物联网安全等
侯嘉慧，中国科学技术大学特任教授。毕业于中国科学技术大学，获得计算机学院学士学位，在美国伊利诺伊理工大学获得博士学位，后于加拿大滑铁卢大学任博士后研究员。2022年加入中国科学技术大学计算机学院。目前主要的研究方向是人工智能安全，多模态学习等。在ACM MobiCom, USENIX Security, IEEE TIFS, IEEE TDSC, IEEE TMC, IEEE INFOCOM等计算机网络及网络与信息安全的一流国际会议及期刊上发表论文三十余篇。
招生信息：欢迎对人工智能安全，多模态学习感兴趣的同学跟我联系。
近年来发表十篇代表性论文(*表示通讯作者)：
Wang Z, Hou J*, Sun H, et al. Task-Oriented Training Data Privacy Protection for Cloud-based Model Training. [C]//34th USENIX Security Symposium. 2025.
2. Zhang J, Liu S, Hou J*, etal. SpeechGuard: Recoverable and Customizable Speech Privacy Protection. [C]//34th USENIX Security Symposium. 2025.
3. Zhang B, Zhang J, Hou J*, et al.TensAllo: Adaptive Deployment of LLMs on Resource-Constrained Heterogeneous Edge Devices. [C]// IEEE International Conference on Computer Communications 2025, London, UK.
4. Zhang J, Hou J*, Tian Ye, et al. WordWhisper: Exploiting Real-Time, Hardware-Dependent IoT Communication Against Eavesdropping. [J]// IEEE Transactions on Mobile Computing.24(1): 15-29 (2025)
5. Luo P, Hou J, Yuan M, et al. F2Zip: Finetuning-Free Model Compression for Scenario-Adaptive Embedded Vision. [C]// In Proceedings of the 22nd ACM Conference on Embedded Networked Sensor Systems,2024, New York, NY, USA, 15–27.
6. Hou J, Liu D, Huang C, et al. Data protection: Privacy-preserving data collection with validation. [J]//IEEE Transactions on Dependable and Secure Computing, 2023, 21.4: 3422-3438.
7. Hou J, Liu H, Liu Y, et al. Model Protection: Real-time Privacy-preserving Inference Service for Model Privacy at the Edge. [J]// IEEE Transactions on Dependable and Secure Computing, 2022, 4270-4284;
8. Hou J, Li X-Y, Zhu P, et al. SignSpeaker: A Real-time, High-Precision SmartWatch-based Sign Language Translator. [C]// International Conference on Mobile Computing and Networking, 2019, 1-15;
9. Hou J, Qian J, Wang Y, et al, ML defense: Against Prediction API Threats in Cloud-Based Machine Learning Service. [C]// IEEE/ACM International Symposium on Quality of Service, 2019, 1-10;
10. Hou J, Li X-Y, Jung T, et al, CASTLE: Enhancing the Utility of Inequality Query Auditing Without Denial Threats. [J]// IEEE Transactions on Information Forensics and Security, 2018, 1656 – 1669;
（更新于2025年4月）