CV

Professional Experience

Researcher, State Key Laboratory of Blockchain and Data Security, Zhejiang University 2024-06 – Present

Member of the AI Data Security Team

• Conducting research on AI-generated content (AIGC) security

• Developing the DFscan platform for multimodal deepfake detection; leading the technical team in platform design and development

Qiushi Research Fellow, ZJU-Hangzhou Global Scientific and Technological Innovation Center 2023-12 – 2024-05

Member of the Cyberspace Security Research Institute

• Conducted research on multimodal data security and privacy protection technologies

• Developed multimodal deepfake detection systems

Postdoctoral Researcher, Zhejiang University 2021-01 – 2023-11

College of Computer Science and Technology

• Researched voice security and privacy protection in human-computer interaction scenarios

• Postdoctoral supervisor: Prof. Kui Ren (Qiushi Chair Professor, AAAS/ACM/CCF/IEEE Fellow, Dean of the College of Computer Science and Technology of Zhejiang University)

Education

Ph.D., Computer Science — Lancaster University, Lancaster, UK

2016-10 – 2020-12

School of Computing and Communications. Thesis: Acoustic-Channel Attack and Defence Methods for Personal Voice Assistants. Supervisors: Prof. Utz Roedig (University College Cork) and Prof. Jeff Yan (University of Southampton)

Dual Master’s Degree, Electrical Engineering (KU Leuven) & Integrated Circuit Engineering (Tsinghua) — KU Leuven & Tsinghua University, Leuven, Belgium & Beijing, China

2012-09 – 2015-09

Department of Electrical Engineering, KU Leuven & School of Integrated Circuits, Tsinghua University. Supervisor: Prof. Guoqiang Bai (Tsinghua University)

B.Eng., Electronic Science and Technology — Beijing University of Posts and Telecommunications, Beijing, China

2008-09 – 2012-06

School of Electronic Engineering

Publications

Representative Publications

• UniAP: Protecting Speech Privacy With Non-Targeted Universal Adversarial Perturbations — IEEE Transactions on Dependable and Secure Computing (2024)

• ALIF: Low-Cost Adversarial Audio Attacks on Black-Box Speech Platforms Using Linguistic Features — 2024 IEEE Symposium on Security and Privacy (SP) (2024)

• SurrogatePrompt: Bypassing the Safety Filter of Text-to-Image Models via Substitution — Proceedings of the ACM SIGSAC Conference on Computer and Communications Security (CCS 2024) (2024)

• Transferring Audio Deepfake Detection Capability Across Languages — Proceedings of the ACM Web Conference 2023 (2023)

• Personal Voice Assistant Security and Privacy—A Survey — Proceedings of the IEEE (2022)

Other Publications (Reverse Chronological Order)

• Beyond Content: A Comprehensive Speech Toxicity Dataset and Detection Framework Incorporating Paralinguistic Cues — Proceedings of the AAAI Conference on Artificial Intelligence (2026)

• Attack-Resistant Watermarking for AIGC Image Forensics via Diffusion-based Semantic Deflection — The 14th International Conference on Learning Representations (ICLR 2026) (2026)

• MixFake: Benchmarking and Enhancing Audio Deepfake Detection in Diverse Real-world Mixed Audio — IEEE International Conference on Multimedia and Expo (ICME 2026) (2026)

• HyperPotter: Spell the Charm of High-Order Interactions in Audio Deepfake Detection — International Conference on Machine Learning (ICML 2026) (2026)

• Spoofed Speech Detection in Real-World Fraudulent Communication Scenarios — Journal of Cyber Security (2025)

• Phoneme-Based Proactive Anti-Eavesdropping With Controlled Recording Privilege — IEEE Transactions on Dependable and Secure Computing (2025)

• SecHeadset: A Practical Privacy Protection System for Real-time Voice Communication — Proceedings of the 23rd Annual International Conference on Mobile Systems, Applications and Services (MobiSys 2025) (2025)

• WMCopier: Forging Invisible Image Watermarks on Arbitrary Images — The 39th Conference on Neural Information Processing Systems (NeurIPS 2025) (2025)

• Deepfake Detection: Key Challenges and Technical Approaches — Computing Magazine of the CCF (2025)

• Indelible “Footprints” of Inaudible Command Injection — IEEE Transactions on Information Forensics and Security (2024)

• InfoMasker: Preventing Eavesdropping Using Phoneme-Based Noise — Proceedings 2023 Network and Distributed System Security Symposium (2023)

• Adversarial Command Detection Using Parallel Speech Recognition Systems — Lecture Notes in Computer Science (2022)

• SonarSnoop: Active Acoustic Side-Channel Attacks — International Journal of Information Security (2020)

• Smart Speaker Privacy Control—Acoustic Tagging for Personal Voice Assistants — 2019 IEEE Security and Privacy Workshops (SPW) (2019)

• Towards Reactive Acoustic Jamming for Personal Voice Assistants — Proceedings of the 2nd International Workshop on Multimedia Privacy and Security (2018)

Preprints / Under Review

• Divide and Conquer: Multimodal Video Deepfake Detection via Cross-Modal Fusion and Localization — arXiv preprint arXiv:2602.00209 (2026)

• JudgeRail: Harnessing Open-Source LLMs for Fast Harmful Text Detection with Judicial Prompting and Logit Rectification — Under review (2025)

• Test-Time Adaptation for Audio Deepfake Detection — Under review (2025)

• Robust Watermarks Leak: Channel-Aware Feature Extraction Enables Adversarial Watermark Manipulation — arXiv preprint arXiv:2502.06418 (2025)

• Bridging the Synthesis-Detection Gap: A Chinese Audio Deepfake Dataset and Industrial-Scale Retrieval-Augmented Detection — Under review (2025)

• CLINDA: A Cross-lingual Domain Adaptation Framework for Challenging Audio Deepfake Detection Tasks across Languages — Under review (2024)

• Masked Diffusion Models Are Fast Distribution Learners — arXiv preprint arXiv:2306.11363 (2023)

Thesis

• Acoustic-Channel Attack and Defence Methods for Personal Voice Assistants — Ph.D. Dissertation, Lancaster University (2020)

Research Projects

• Research on Speech Synthesis Data Compliance Management Technology Based on Intrinsic Characteristics of Audio Signals (Principal Investigator) — National Natural Science Foundation of China (NSFC) (2025–2028)

• High-Performance Visual Perception Models Using Deep Learning (Participant) — National Natural Science Foundation of China (NSFC) (2023–2026)

• Cross-Chain Security in Heterogeneous Blockchain Networks (Participant) — National Natural Science Foundation of China (NSFC) (2023–2027)

• Research on Voice Attack and Defense Based on the Physical Characteristics of Smart Device Sensing Components (Participant) — National Natural Science Foundation of China (NSFC) (2022–2025)

• Multimodal Network Environment Construction Technology Based on Public Cloud-Network Resources (Participant) — Key R&D Programs of China (2024–2027)

• Aggregation and Transfer of Machine Learning Models (National Science and Technology Innovation 2030 Initiative - New Generation Artificial Intelligence) (Participant) — Key R&D Programs of China (2021–2025)

• Security Protection Technology for Industrial Control Programming Platforms Based on Domestic Cryptographic Algorithms (Participant) — Key R&D Programs of China (2022–2024)

• Key R&D Programme of Zhejiang Province (Participant) — Provincial, Municipal, and University-Level Projects (2024–2026)

• Key Technologies and Platform Development for Security Detection of Large AI Models (Participant) — Hangzhou Key R&D Program (2024–2027)

• Toolchain for Deep Synthetic Content Analysis Based on Physical Attribute Attribution (Principal Investigator) — Hangzhou West Innovation Corridor Development Special Fund (2024–2026)

• Active and Passive Security Protection Technologies for the Maojing Voice Interaction System (Principal Investigator) — Zhejiang University-Alibaba Collaboration Project (2025–2026)

• Security Risk Detection and Alignment Strategies for Large Model-based AI Agents (Participant) — Zhejiang University-Ant Group Joint Laboratory (2025–2025)

• Research and Development of AI-Driven Automated Security Detection Technologies and Components for Power Systems (Participant) — China Southern Power Grid Research Institute Co., Ltd. (2024–2026)

• Deep Learning Algorithm Evaluation and Security Verification Platform (Participant) — CRRC Zhuzhou Electric Locomotive Research Institute Co., Ltd. (2024–2027)

• Research on AI Attack-Defense Library Design and Test Component Development for New Power System Scenarios (2023) (Participant) — China Southern Power Grid Research Institute Co., Ltd. (2023–2025)

• Multidimensional Protocol Carrying and Endogenous Security Access Technologies for Multimodal Networks (Participant) — Key R&D Programs of China (2024–2027)

Honors & Awards

• Silver Award — IJCAI 2026 Deepfake Detection and Localization Challenge (DDL 2.0) (2026-08) — Team AIGVDete (Zhejiang University)

• ICML 2026 Silver Reviewer — International Conference on Machine Learning (ICML 2026) (2026-05) — Top-tier Silver Reviewer distinction awarded by the ICML 2026 Program Chairs

• 3rd Place Award — Workshop & Challenge on Deepfake Detection, Localization, and Interpretability, IJCAI 2025 (2025-08) — Advisor — Track: Audio-Video Detection and Localization (DDL-AV)

• National Grand Prize (Top-Tier Award) — 19th “Challenge Cup” National Competition for Extracurricular Academic Science and Technology Works (2024-11) — Advisor (Ranked 2nd among 3 advisors) — Project: “Multimodal AI Audit Matrix: Deepfake Detection and NSFW Content Regulation Platform”

• National Graduate Scholarship Achievement (October 2024) (2024-10) — Co-supervised Master’s student won the National Graduate Scholarship (China)

• Top 5 Nationwide — 3rd China Artificial Intelligence Competition (2021-12) — Primary Advisor (Ranked 1st among 2 advisors) — Audio Deepfake Detection Under Open-Speaker Scenarios; Speaker-Specific Audio Deepfake Detection

• Postdoctoral Excellence Grant (Second Class) — Zhejiang Provincial Department of Human Resources and Social Security (2021-08) —

• Finalist for “Most Innovative Research” Pwnie Award — Black Hat USA (2019-08) — First Author (1st of 4 contributors) — Research: “SonarSnoop: Active Acoustic Side-Channel Attacks”

• Ph.D. Scholarship — Faculty of Science and Technology, Lancaster University, UK (2016-10) —

Student Supervision & Mentorship

• Co-supervising 5 Ph.D. students, 7 Master’s students and one undergraduate student at Zhejiang University on research projects in AIGC security and multimodal privacy

• Co-supervised a Zhejiang University Master’s student to win the National Graduate Scholarship (China) — the highest-level scholarship awarded to Master’s students in China (Oct 2024)

• Successfully mentored 1 PhD student, 5 Master’s students, and 5 undergraduates to degree completion, guiding thesis design, research execution, and publication strategies

• Teaching Assistant for lab/practical sessions: CS4615: Computer Systems Security, University College Cork, Ireland; SCC110: Software Development, Lancaster University, UK

Academic Services

Journal Editorial Roles

• Special Issue Initiator and Guest Editor-in-Chief: “Intelligent Voice Security and Defense Technologies”, Journal of Cyber Security, 2025

Conference Program Committees

• USENIX Security 2027, AAAI 2027

Conference Reviewer

• ICLR 2027, Interspeech 2026, AAAI 2026, IEEE SLT 2026, ACM MM 2026, ICLR 2026, ICML 2026, ACM Web Conference (WWW 2025)

Journal Reviewer

• English Journals: Proceedings of the IEEE, IEEE Transactions on Information Forensics and Security (TIFS), IEEE Transactions on Dependable and Secure Computing (TDSC), IEEE Internet of Things Journal (IoT-J), ACM Transactions on Internet of Things (TIOT)

• Chinese Journal: Journal of Information Network Security

Reviewer Recognition

• ICML 2026 Silver Reviewer

Other Academic Activities

• Contributed to the preparation of a €720,839 grant proposal (Science Foundation Ireland – SFI) on Security and Privacy of Personal Voice Assistants under Prof. Utz Roedig (2019)

• Visiting Scholar at the Department of Computer Science, University College Cork (UCC), Ireland (2019–2020)

• Participated in the São Paulo Advanced Science School (ESPCA) on Smart Cities, hosted by the University of São Paulo (2017); selected as one of 75 global top graduate students and postdoctoral researchers (sponsored by FAPESP)

Impacts

Industry Contributions

• Proposed SurrogatePrompt, a prompt attack method to bypass safeguards of commercial large-scale AI models (e.g., Midjourney, Stability.ai), enabling systematic generation of policy-violating content. Findings were acknowledged by these leading vendors.

• Developed WMCopier, an adversarial watermark forgery method that compromises commercial watermarking systems; recognized by Amazon’s Responsible AI Team for identifying critical vulnerabilities (official thank-you letter).

• Developed and open-sourced ALIF, the first black-box adversarial attack framework leveraging linguistic features; adopted by NVIDIA for integration into their official AI security toolkit.

Media Recognition

• SonarSnoop: Active Acoustic Side-Channel Attacks — garnered attention from IT media (Motherboard, ZDNet, Sophos); praised by Bruce Schneier and Prof. Ross Anderson.