|
|
Join Us!
I am looking for highly motivated and self-driven students. If you are interested in intelligent security and privacy techniques, send an email for Ph.D. or Master positions!
About Me
I received my Bachelor degree in computer science from Nanjing University in Jul. 2016. I obtained my Ph.D. degree under the supervision of Prof. Fengyuan Xu at Nanjing University in Dec. 2021. From Feb. 2022, I have joined the COSEC Research Center of NJU and the National Key Lab for Novel Software Technology.
My research interests primarily include LLM Security and Trustworthy Multimodal Agent.
Preprints
- Stop Wandering: Efficient Vision-Language Navigation via Metacognitive Reasoning. Paper
- Em-Garde: A Propose-Match Framework for Proactive Streaming Video Understanding. Paper slides
- From Transactions to Exploits: Automated PoC Synthesis for Real-World DeFi Attacks. Paper
- Zero-Permission Manipulation: Can We Trust Large Multimodal Model Powered GUI Agents? Paper
- Discovering 100+ Compiler Defects in 72 Hours via LLM-Driven Semantic Logic Recomposition. Paper
- A Systematic Study of Code Obfuscation Against LLM-based Vulnerability Detection. Paper
- Revealing Adversarial Smart Contracts through Semantic Interpretation and Uncertainty Estimation. Paper
- Video-in-the-Loop: Span-Grounded Long Video QA with Interleaved Reasoning. Paper
- AdaNav: Adaptive Reasoning with Uncertainty for Vision-Language Navigation. Paper
- When LLMs Copy to Think: Uncovering Copy-Guided Attacks in Reasoning LLMs. Paper(Workshop Paper)
- Everything You Wanted to Know About LLM-based Vulnerability Detection But Were Afraid to Ask. paper
- If LLMs Would Just Look: Simple Line-by-line Checking Improves Vulnerability Localization. paper
- “MCP Does Not Stand for Misuse Cryptography Protocol”: Uncovering Cryptographic Misuse in Model Context Protocol at Scale. Paper
Selected Publications (chronological order)
2026
- [ICDCS 2026] Lingzi Zhao, Huali Lu, Hao Wu, Shucheng Li, Longye Li, Tao He, Wenlong Liao, Feng Lyu. Integrated Load-Balanced Scheduling for Human-Vehicle Collaborative Urban Sanitation.
- [ISSTA 2026] Zhengyang Shan, Xu Qian, Jiayun Xin, Minghui Xu, Yue Zhang, Zhen Yang, Hao Wu, Xiuzhen Cheng. SAGE: Signal-Amplified Guided Embeddings for Vulnerability Detection.
- [SIGIR 2026] Fan Wu, Haoye Pan, Hao Wu, Kai Qian, Shucheng Li, Feng Lyu. SynDiSC: High-Quality Tabular Data Synthesis with Distributional and Semantic Consistency.
- [SIGIR 2026] Shucheng Li, Weixuan Xu, Le Jiang, Hao Wu, Fengyuan Xu, Fan Wu, Feng Lyu. Seeing the Whole Through the Parts: Discovering Objects through Semantic Part Mining in Weak Supervision.
- [OSDI 2026] Jun Xiao, Qinhui Gu, Ligeng Chen, Lizhi Sun, Zicheng Wang, Yinggang Guo, Lu Liu, Hao Wu, Borui Li. Surviving the Impossible Trinity: Revisiting CPU Scheduling Problem on Modern COTS Mobile Devices (Operational Systems).
- [ICME 2026] Wang Yang, Chen Mao, Ligeng Chen, Mingzhe Gao, Shucheng Li, and Hao Wu. Extend Multi-Scale Subgraph Features to Enhance Node Representation for Robust Graph Matching.
- [ToN 2026] Jie Zhao, Feng Lyu, Hao Wu, Fan Wu, Shucheng Li. Game in Motion: Heterogeneous Task Offloading in Dynamic Vehicular Edge Computing.
- [INFOCOM 2026] Qian Kai, Feng Lyu, Hao Wu, Jing Gao, Shucheng Li, Fan Wu, Qiong Luo, Bowen Chen, Fengyuan Xu. KAT: Knowledge-Context Augmentation for Evolving LLM-Based Telecom Troubleshooting.
- [MobiCom 2026] Qiyong Fu, Feng Lyu, Mingliu Liu, Hao Wu, Jieyu Zhou, Lijuan He, Fan Wu, Jinli Sun, Manjia Liu. A3TP: Automated, Accurate, and Adaptive UAV Task Planning for Large-Scale Power Transmission Networks Inspection.
- [AAAI 2026] Wendi Li, Hao Wu, Han Gao, Bing Mao, Fengyuan Xu, and Sheng Zhong. Diverse Human Driving Vehicle Simulation in Background Traffic for Autonomous Driving Tests.
2025
- [TMC 2025] Fei Zeng, Feng Lyu, Hao Wu, Zhanxi Li, Shucheng Li, and Fengyuan Xu. H2O: Heterogeneity-Aware Hierarchical Orchestration for Memory-Efficient On-Device LLM Inference.
- [ASE 2025] Yi Qian, Fei Peng, Hao Wu, Ligeng Chen, and Bing Mao. Uncovering Prompt Elements: Cloning System Prompts from Behavioral Traces.
- [TIFS 2025] Xin Zhao, Hao Han, Hao Wu, Sheng Zhong, and Fengyuan Xu. UTRDCL: Stealthy DCL-Based Obfuscation and Its Attacks and Defenses in Android.
- [MM 2025] Yixin Xu, Hao Wu, Jingzhou Zhu, Fengyuan Xu, Sheng Zhong. PriCAF: Privacy-Preserving Contribution Assessment in Federated Learning Before Model Training.
- [ICCV 2025] Xin Ding, Hao Wu, Yifan Yang, Shiqi Jiang, Donglin Bai, Zhibo Chen, and Ting Cao. StreamMind: Unlocking Full Frame Rate Streaming Video Dialogue through Event-Gated Cognition. paper
- [MobiCom 2025] Feng Lyu, Lijuan He, Mingliu Liu, Sijing Duan, Hao Wu, Jieyu Zhou, Yi Ding, Zaixun Ling, and Yibo Cui. Auto-UIT: Automated UAV Inspection Trajectory Generation from Noisy Sparse 3D Point Cloud.
- [KDD 2025] Likun Zhang, Hao Wu, Lingcui Zhang, Fengyuan Xu, Jin Cao, Fenghua Li, Ben Niu. Training Data Attribution: Was Your Model Secretly Trained On Data Created By Mine? paper
- [FSE 2025] Xing Su, Hanzhong Liang, Hao Wu, Ben Niu, Fengyuan Xu, and Sheng Zhong. DiSCo: Towards Decompiling EVM Bytecode to Source Code using Large Language Models.
- [Oakland 2025] Xiao Li, Yue Li, Hao Wu, Yue Zhang, Kaidi Xu, Xiuzhen Cheng, Sheng Zhong, and Fengyuan Xu. Make a Feint to the East While Attacking in the West: Blinding LLM-Based Code Auditors with Flashboom Attacks.
2024
- [TrustCom 2024] Lizhi Sun, Jingzhou Zhu, Boyu Chang, Yixin Xu, Bo Yang, Hao Wu, Fengyuan Xu, and Sheng Zhong. TTFL: Towards Trustworthy Federated Learning with Arm Confidential Computing.
- Shimao Xu, Xiaopeng ke, and Hao Wu. CAPter: Controllable Data Privacy Enhancement for Deep Learning Inference Services. UIC 2024.
- [TKDD 2024] Shucheng Li, Jingzhou Zhu, Boyu Chang, Hao Wu, Fengyuan Xu, and Sheng Zhong. Multi-Label and Evolvable Dataset Preparation for Web-Based Object Detection.
- [TIFS 2024] Hao Wu, Yuhang Gong, Xiaopeng Ke, Hanzhong Liang, Fengyuan Xu, Yunxin Liu, and Sheng Zhong. TIM: Enabling Large-scale White-box Testing on In-App Deep Learning Models.
- [MM 2024] Hao Wu, Likun Zhang, Shucheng Li, Fengyuan Xu, and Sheng Zhong. CoAst: Validation-Free Contribution Assessment for Federated Learning based on Cross-Round Valuation. paper
- [ICRA 2024] Han Gao, Yating Liu, Fang Cao, Hao Wu, Fengyuan Xu, and Sheng Zhong. VIDAR: Data Quality Improvement for Monocular 3D Reconstruction through In-situ Visual Interaction. paper
- [ICASSP 2024] Han Gao, Hao Wu, Peiwen Dong, Yixin Xu, Fengyuan Xu, and Sheng Zhong. MuSR: Multi-Scale 3D Scenes Reconstruction based on Monocular Video. paper
2023
- [MM 2023] Shucheng Li, Runchuan Wang, Hao Wu, Sheng Zhong, and Fengyuan Xu. SIEGE: Self-Supervised Incremental Deep Graph Learning for Ethereum Phishing Scam Detection. paper
- [SIGIR 2023] Shucheng Li, Boyu Chang, Bo Yang, Hao Wu, Sheng Zhong, and Fengyuan Xu. Dataset Preparation for Arbitrary Object Detection: An Automatic Approach based on Web Information in English. paper
- [ICASSP 2023, Oral] Hao Wu, Bo Yang, Xiaopeng Ke, Siyi He, Fengyuan Xu, and Sheng Zhong. GAPter: Gray-box Data Protector for Deep Learning Inference Services at User Side. paper
2022
- [TMC 2022] Lizhi Sun, Shuocheng Wang, Hao Wu, Yuhang Gong, Fengyuan Xu, Yunxin Liu, Hao Han, and Sheng Zhong. LEAP: TrustZone Based Developer-Friendly TEE for Intelligent Mobile Apps. paper
- [TrustCom 2022] Ligeng Chen, Zhongling He, Hao Wu, Yuhang Gong, and Bing Mao. AVMiner: Expansible and Semantic-Preserving Anti-Virus Labels Mining Method. paper
- [ICASSP 2022] Xiaopeng Ke, Boyu Chang, Hao Wu, Fengyuan Xu, and Sheng Zhong. Towards Practical and Efficient Long Video Summary. paper
- [计算机研究与发展 2022] 吴昊, 王浩, 苏醒, 李明昊, 许封元, 仲盛. 自动驾驶系统中视觉感知模块的安全测试. paper
- [SANER 2022] Ligeng Chen, Zhongling He, Hao Wu, Fengyuan Xu, Yi Qian, and Bing Mao. DIComP: Lightweight Data-Driven Inference of Binary Compuler Provenance with High Accuracy.Paper
2021
- [WWW 2021] Hao Wu, Xuejin Tian, Yuhang Gong, Xing Su, Minghao Li, and Fengyuan Xu. DAPter: Preventing User Data Abuse in Deep Learning Inference Services. paper slides
- [MobiCom 2021] Hao Wu, Xuejin Tian, Minghao Li, Yunxin Liu, Ganesh Ananthanarayanan, Fengyuan Xu, and Sheng Zhong. PECAM: Privacy-Enhanced Video Streaming & Analytics via Securely-Recoverable Transformation. paper slides
2020
- [MobiSys 2020] Hao Wu, Jinghao Feng, Xuejin Tian, Edward Sun, Yunxin Liu, Bo Dong, Fengyuan Xu, and Sheng Zhong. EMO: Real-Time Emotion Recognition from Single-Eye Images for Resource-Constrained Eyewear Devices. paper slides
Awards
- 中央某部委高层次创新人才
- 南京大学首届“苗圃计划” 2025
- MSR Asia StarTrack Scholars 2024 program
- 中文信息学会 优秀博士学位论文
- ACM SIGBED（中国）优博
- 江苏省计算机学会优秀博士学位论文
- 南京大学计算机科学与技术系优秀博士论文
- “紫金山英才”栖霞先锋计划高层次创新创业人才
Fundings
- 2025-2029, NSFC重点项目, 任务负责人
- 2024-2026, NSFC青年科学基金, 云边端融合下隐私增强的高可用智能计算协同技术，主持
- 2023-2024, NSFC基金，任务负责人
- 2022-2025, 国家重点研发计划项目，子课题负责人
- 2022,CCF-华为胡杨林基金, 面向隐私计算的边缘协同训练新范式研究，主持
- 2022-2025, 江苏省青年基金, 端边协同的移动智能计算中的支撑技术研究，主持
Academic Services
- 中文信息学会 大数据安全与隐私计算专委会 委员
- 中国网络空间安全学会 宣传工作委员会 委员
- ACM SIGBED中国 委员