권태경 교수
- 02-2123-4523
- taekyoung@yonsei.ac.kr
- 정보보호연구실/AI 보안연구실
- 새천년관 407호
주요링크
연구분야
AI SecurityAI SafetyData PrivacySystem Security
학력
1995-1999 : 연세대학교 컴퓨터과학과 박사
1993-1995 : 연세대학교 컴퓨터과학과 석사
1988-1992 : 연세대학교 컴퓨터과학과 학사
경력
2013-현재 : 연세대학교 정보대학원 교수
2020-현재 : 연세대학교 인공지능대학원 AI+X 겸직교수
2001-2013 : 세종대학교 컴퓨터공학과 교수
1999-2000 : Univ. of California at Berkeley, Postdoctoral Researcher
주요 연구활동 및 저서
-
Amplifying Training Data Exposure through Fine-Tuning with Pseudo-Labeled Memberships, IEEE Transactions on Information Forensics and Security (TIFS), 2025. (IF: 8)
-
Red-Teaming LLMs with Token Control Score: Efficient, Universal, and Transferable Jailbreaks, RAID 2025. (BK: 2)
-
Multi-View Slot Attention Using Paraphrased Texts for Face Anti-Spoofing, ICCV 2025. (BK: 4)
-
Continuous Authentication for Secure and Seamless User-Avatar Integration in Multi-Device Metaverses, IEEE Internet of Things Journal (IoT-J), 2025. (IF: 8.9)
-
A Continuous Authentication Framework for Securing Metaverse Identities, IEEE Transactions on Services Computing (TSC), 2025. (IF: 5.8)
-
Fuzzing JavaScript Interpreters with Coverage-Guided Reinforcement Learning for LLM-based Mutation, in Proc. the 33rd ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA), 2024. (BK: 2)
-
Adversarial Feature Alignment: Balancing Robustness and Accuracy in Deep Learning via Adversarial Training, AIsec 2024
-
Enhancing Differential Fuzzing of Cryptographic Libraries with Sustainable Hybrid Fuzzing and Crypto-Specific Mutation, ICISC 2024 (Best Paper Award)
-
BoKASAN: Binary-only Kernel Address Sanitizer for Effective Kernel Fuzzing, Usenix Security 2023 (BK: 3)