Posts by Collection

portfolio

Portfolio item number 1

Short description of portfolio item number 1

Portfolio item number 2

Short description of portfolio item number 2

publications

A Dual Stealthy Backdoor: From Both Spatial and Frequency Perspectives

Published in AAAI Conference on Artificial Intelligence (AAAI), 2024

A stealthy backdoor designed from both spatial and frequency perspectives. (CCF-A)

Recommended citation: Y. Gao, H. Chen, P. Sun, J. Li, A. Zhang, Z. Wang, W. Liu. (2024). "A Dual Stealthy Backdoor: From Both Spatial and Frequency Perspectives." AAAI Conference on Artificial Intelligence (AAAI). (CCF-A)
Download Paper

Investigating the Backdoor on DNNs Based on Recolorization and Reconstruction: From a Multi-Channel Perspective

Published in IEEE Transactions on Information Forensics and Security (TIFS), 2024

Investigating DNN backdoors based on recolorization and reconstruction. (CCF-A)

Recommended citation: H. Chen, Y. Gao, A. Zhang, P. Sun, N. Jiang, W. Liu, X. Wang. (2024). "Investigating the Backdoor on DNNs Based on Recolorization and Reconstruction." IEEE Transactions on Information Forensics and Security (TIFS). (CCF-A)
Download Paper

Energy-Based Backdoor Defense Without Task-Specific Samples and Model Retraining

Published in International Conference on Machine Learning (ICML), 2024

An energy-based backdoor defense requiring no task-specific samples or model retraining. (CCF-A)

Recommended citation: Y. Gao, H. Chen, P. Sun, Z. Li, J. Li, H. Shao. (2024). "Energy-Based Backdoor Defense Without Task-Specific Samples and Model Retraining." International Conference on Machine Learning (ICML). (CCF-A)
Download Paper

Lightweight Privacy Protection Against Channel-Level Gradient Leakage Attack in Federated Learning

Published in International Conference on Information and Communication Technology Convergence (ICTC), 2024

Lightweight privacy protection against channel-level gradient leakage in federated learning. Best Paper Award.

Recommended citation: Z. Li, H. Chen, Y. Gao, H. Xue. (2024). "Lightweight Privacy Protection Against Channel-Level Gradient Leakage Attack in Federated Learning." International Conference on Information and Communication Technology Convergence (ICTC). Best Paper Award.
Download Paper

Frequency Is Devious: Deep Stealthy Backdoor in Artificial Intelligence-Based Consumer Internet of Things

Published in IEEE Transactions on Consumer Electronics (TCE), 2025

A deep stealthy frequency-domain backdoor in AI-based consumer IoT.

Recommended citation: Y. Gao, H. Chen, J. Li, X. Wang, A. Alharbi, A. Tolba, F. Xia. (2025). "Frequency Is Devious: Deep Stealthy Backdoor in Artificial Intelligence-Based Consumer Internet of Things." IEEE Transactions on Consumer Electronics (TCE).
Download Paper

A Triple Stealthy Backdoor: Hidden in Spatial, Frequency, and Feature Domains

Published in IEEE Transactions on Dependable and Secure Computing (TDSC), 2025

A stealthy backdoor hidden across spatial, frequency, and feature domains. (CCF-A)

Recommended citation: Y. Gao, H. Chen, P. Sun, J. Li, Y. Yin, Z. Wang, W. Liu. (2025). "A Triple Stealthy Backdoor: Hidden in Spatial, Frequency, and Feature Domains." IEEE Transactions on Dependable and Secure Computing (TDSC). (CCF-A)
Download Paper

SkillReducer: Optimizing LLM Agent Skills for Token Efficiency

Published in arXiv preprint, 2026

Optimizing LLM agent skills for token efficiency.

Recommended citation: Y. Gao, Z. Li, Y. Yuan, Z. Ji, P. Ma, S. Wang. (2026). "SkillReducer: Optimizing LLM Agent Skills for Token Efficiency." arXiv preprint.
Download Paper

CM-EVS: Sparse Panoramic RGB-D-Pose Data for Complete Scene Coverage

Published in Conference on Neural Information Processing Systems (NeurIPS), 2026

A coverage-curated panoramic RGB-D-pose dataset: 36,373 ERP frames from 1,275 indoor scenes, with a median of only 25 frames per scene. Corresponding author.

Recommended citation: J. Liu, J. Li, J. Yu, X. Yu, et al., Y. Gao (corresponding author), M. Cheng. (2026). "CM-EVS: Sparse Panoramic RGB-D-Pose Data for Complete Scene Coverage." NeurIPS. Accepted. (CCF-A)
Download Paper

One Cognitive Loop Is Enough: SODA unlocks Pure-Text Spatial Reasoning in Large Language Models

Published in Annual Meeting of the Association for Computational Linguistics (ACL), 2026

A single cognitive loop unlocks pure-text spatial reasoning in LLMs. Corresponding author. (CCF-A)

Recommended citation: S. Bai, J. Zhang, H. Huang, Y. Wang, J. Liu, Y. Wu, N. Yu, Y. Gao (corresponding author), M. Cheng. (2026). "One Cognitive Loop Is Enough: SODA unlocks Pure-Text Spatial Reasoning in Large Language Models." Annual Meeting of the Association for Computational Linguistics (ACL). (CCF-A)
Download Paper

Bit-Flip Attacks on Vision-Language-Action Models: Action-Decoding Architecture Shapes the Vulnerability

Published in arXiv preprint, 2026

Evaluating weight-integrity vulnerabilities and localized protection across vision-language-action architectures.

Recommended citation: Y. Gao, L. Chen, W. Wu, M. Zhou, J. Wang, K. Ji, M. Guo, H. Chen. (2026). "Bit-Flip Attacks on Vision-Language-Action Models: Action-Decoding Architecture Shapes the Vulnerability." arXiv:2608.15475.
Download Paper

PathMark: Protecting Intellectual Property of Mixture-of-Expert LLMs via Path Watermarks

Published in ACM Conference on Computer and Communications Security (CCS), 2026

Path watermarks for protecting the intellectual property of Mixture-of-Expert LLMs. (CCF-A)

Recommended citation: Y. Gao, Q. Wang, Y. Yuan, R. Huang, L. Chen, Z. Ji, S. Wang. (2026). "PathMark: Protecting Intellectual Property of Mixture-of-Expert LLMs via Path Watermarks." ACM Conference on Computer and Communications Security (CCS). (CCF-A)
Download Paper

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015