Xinghao Yang （杨兴浩）
Home
 |
Xinghao Yang
Associate Professor
|
Brief Biography
-
Xinghao Yang is currently a Associate Professor with the College of Control Science and Engineering, China University of Petroleum (East China), China. He received the BEng degree in electronic information engineering and MEng degree in information and communication engineering from China University of Petroleum (East China) in 2015 and 2018, respectively. He received the Ph.D. degree in computer science from the University of Technology Sydney in 2022. His current research interests include natural language processing and AI security. My work focuses on attack-defense mechanisms and robustness protection of large models and intelligent systems, covering jailbreaking attacks and defenses on large models, textual adversarial attacks and defenses, textual backdoor attacks and defenses, multimodal model security, as well as security of large reasoning models and RAG systems.
Research
My research interests lie in machine learning and its real-world applications, covering multiview learning, manifold learning and deep learning.
My current research interest: AI Robustness.
Selected Publications
Zikai Wang, Xinghao Yang, Baodi Liu, Wei Liu, Weifeng Liu. Discrete multi-head training with rule-based perturbation location for accurate text defense. Pattern Recognition, 2026, 115092.
Bin Xu, Yinghua Li, Linqi Han, Xiaoyu Li, Xinghao Yang, Wei Liu, Yongshun Gong. IMTS-Tokenizer: Time-Aware Tokenization for Irregular Multivariate Time Series Forecasting. Advances in Neural Information Processing Systems (NeurIPS 2026)., December 6-12, 2026, Sydney, Australia.
Mingze Ni, Yongshun Gong, Xinghao Yang, and Wei Liu. Is LoRA Robust to Adversarial Attacks? Expectation-Based Regularization for Parameter-Efficient Adversarial Training. In Findings of the Association for Computational Linguistics: EMNLP 2026., October 24-29, 2026, Budapest, Hungary.
Puhua Jia, Xinghao Yang, Yongwei Tang, Baodi Liu, Wei Li and Weifeng Liu. Universal Attack based Focal Enhancement for Bearing Fault Diagnosis. Knowledge-Based Systems, 2026, 348: 116344.
Ao Wang, Xinghao Yang, Yongshun Gong, Wei Liu, Baodi Liu and Weifeng Liu. Enhancing the Transferability of Jailbreak Attacks on Large Language Models via Exploiting Reparameterization Invariance. ACL 2026, July 2-7, 2026, San Diego.
Zikai Wang, Xinghao Yang, Bingfeng Zhang, and Weifeng Liu. Safety Anchor-Guided Adaptive Bias Decay for Jailbreak Defense. ICASSP 2026, May 4-8, 2026, Barcelona.
Jiawei Wang, Xinghao Yang, Baodi Liu and Weifeng Liu. Intermediate-Layer Filtering for Text Backdoor Defense. ICNLP 2026, March 20-22, 2026, Xi'an. Best Paper
Chen Li, Xinghao Yang, Ao Wang, Yongshun Gong, Baodi Liu and Weifeng Liu. Multiobjective Simulated Annealing-Based Stopwords Substitution for Rubbish Text Attack. IEEE Transactions on Neural Networks and Learning Systems, 2026.
Yang An, Zhibin Li, Xiaoyu Li, Wei Liu, Xinghao Yang, Haoliang Sun, Meng Chen, Yu Zheng, Yongshun Gong. Spatio-Temporal Multivariate Probabilistic Modeling for Traffic Prediction. IEEE Transactions on Knowledge and Data Engineering, 2025, 37: 2986-3000.
Yidan Xu, Xinghao Yang, Wei Liu, Bao-di Liu, Weifeng Liu. Disentangled Information Bottleneck for Adversarial Text Defense. EMNLP, November 4-9, 2025, Suzhou.
Ao Wang, Xinghao Yang, Chen Li, Bao-di Liu, Weifeng Liu. Adaptive Immune-based Sound-Shape code Substitution for Adversarial Chinese Text Attacks. EMNLP 2024, November 12-16, 2024, Miami.
Jian Pang, Weifeng Liu, Bingfeng Zhang, Xinghao Yang, Baodi Liu, Dapeng Tao. MCNet: Magnitude consistency network for domain adaptive object detection under inclement environments. Pattern Recognition, 2024, 145: 109947.
Chen Li, Xinghao Yang, Baodi Liu, Weifeng Liu, Honglong Chen. Annealing Genetic-based Preposition Substitution for Text Rubbish Example Generation. IJCAI 2023, August 19-25, 2023, Macao.
Renjie Xu, Xinghao Yang, Xingxing Yao, Dapeng Tao, Weijia Cao, Xiaoping Lu, Weifeng Liu. Self-Paced Hard Task-Example Mining for Few-Shot Classification. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33: 5631-5644.
Xinghao Yang, Yongshun Gong, Weifeng Liu, James Bailey, Dacheng Tao, Wei Liu. Semantic-Preserving Adversarial Text Attacks. IEEE Transactions on Sustainable Computing, 2023, 8: 583-595.
Xinghao Yang, Yupeng Qi, Honglong Chen, Baodi Liu, Weifeng Liu. Generation-based parallel particle swarm optimization for adversarial text attacks. Information Sciences, 2023, 644: 119237.
Xinghao Yang, Weifeng Liu, Wei Liu. Tensor Canonical Correlation Analysis Networks for Multi-view Remote Sensing Scene Recognition. IEEE Transactions on Knowledge and Data Engineering, 2022, 34: 2948-2961.
Xinghao Yang, Weifeng Liu, Shengli Zhang, Wei Liu, Dacheng Tao. Targeted Attention Attack on Deep Learning Models in Road Sign Recognition. IEEE Internet of Things Journal, 2021, 8: 4980-4990.
Aneesh Chivukula, Xinghao Yang, Tianqing Zhu, Wanlei Zhou, and Wei Liu. Game Theoretical Adversarial Deep Learning With Variational Adversaries. IEEE Transactions on Knowledge and Data Engineering, 2021, 33: 3568-3581.
Xinghao Yang, Weifeng Liu, Wei Liu, Dacheng Tao. A Survey on Canonical Correlation Analysis. IEEE Transactions on Knowledge and Data Engineering, 2021, 33: 2349-2368.
Xinghao Yang, Weifeng Liu, Dacheng Tao, Wei Liu. BESA: BERT-based Simulated Annealing for Adversarial Text Attacks. IJCAI 2021, August 21-26, 2021.
Xinghao Yang, Weifeng Liu, James Bailey, Dacheng Tao, and Wei Liu. Bigram and Unigram Based Text Attack via Adaptive Monotonic Heuristic Search. AAAI 2021, February 2-9, 2021.
Xinghao Yang and Wei Liu. Population Location and Movement Estimation through Cross-domain Data Analysis. IJCAI 2020, January 7-15, 2021.
Weifeng Liu, Xinghao Yang, Dapeng Tao, Jun Cheng, Yuanyan Tang. Multiview dimension reduction via Hessian multiset canonical correlations. Information Fusion, 2018, 41: 119 - 128.
Xinghao Yang, Weifeng Liu, Dapeng Tao, Jun Cheng. Canonical correlation analysis networks for two-view image recognition. Information Sciences, 2017, 385: 338-352.
Xinghao Yang, Weifeng Liu, Dapeng Tao, Jun Cheng, Shuying Li. Multiview Canonical Correlation Analysis Networks for Remote Sensing Image Recognition. IEEE Geoscience and Remote Sensing Letters, 2017, 14: 1855-1859.
Xinghao Yang, Weifeng Liu. Multiple Scale Canonical Correlation Analysis Networks for Two-View Object Recognition. The 24th International Conference on Neural Information Processing (ICONIP 2017), November 14-18, 2017, Guangzhou, China.
Projects
Young Talent of Lifting engineering for Science and Technology in Shandong, China (Grant No. SDAST2024QTA040), Jan. 2025 - Dec. 2026.
Research on Bounded Text Attack Method based on Monotone Weak Submodular Optimization, the National Natural Science Foundation of China (Grant No. 62406341), Jan. 2025 - Dec. 2027.
Research on Robust Text Data Representation based on Adversarial Attack and Defense, the Shandong Natural Science Foundation (Grant No. ZR2023QF051), Jan. 2024 - Dec. 2026.
Research on Robust Representation Learning Method based on Data Model Coupling, Outstanding Youth Science Foundation Project of Shandong Province (Overseas) (Grant No.2023HWYQ-070), Jan. 2023 - Dec. 2025.
Text Adversarial Research based on Multivariate Combination Optimization, Independent Innovation Research Project (Grant No.22CX06059A), Sep. 2022 - Apr. 2025.