Longwei Wang

Longwei Wang is an Assistant Professor of Computer Science whose research focuses on explainable machine learning and trustworthy AI, particularly the interplay among AI explainability, robustness, and security. His research focuses on two fundamental challenges in modern AI: the black-box nature of deep learning models and the prevalence of hallucinations and adversarial vulnerabilities in machine learning systems.

His research aims to bridge the gap between interpretability and robustness by designing machine learning models that are not only understandable to humans but also resilient to adversarial attacks. Together with his collaborators, he developed Winsor-CAM, an interpretable visual explanation framework for deep neural networks, and has contributed to the theoretical foundations linking explainability with adversarial robustness. His work has appeared in leading venues, including ICML, NeurIPS, IEEE Transactions on Pattern Analysis and Machine Intelligence, ICDM, AAAI, the ACM Web Conference, and numerous IEEE journals. His research has received several recognitions, including the Best Paper Award at ISPR 2025 and a NeurIPS 2025 Spotlight Recognition.

As an active member of the research community, he has served as a Special Track and Workshop Chair for the International Conference on Recent Trends in Image Processing and Pattern Recognition and as a Program Committee member for major AI conferences, including ICML, NeurIPS, ICLR, AAAI, IJCAI, WWW, IEEE CAI, and IEEE CogMI. He also serves as an Editor for the journal Symmetry. Dr. Wang regularly serves as a reviewer for high-impact journals such as ACM Transactions on Intelligent Systems and Technology, IEEE Transactions on Artificial Intelligence, IEEE Transactions on Systems, Man, and Cybernetics. He is an IEEE and ACM member.

CSC 488/588 Pattern Recognition and Machine Learning
CSC 525 High Performance Computing
CSC 790 Seminar: Advanced topics in deep learning
CSC 492/592 Internet of Things

Robust and Explainable AI, Trustworthy Deep Learning, Equivariant Deep Learning, Compositional Generalization, Dynamic Routing, Explainable AI for Medical analysis, and XAI for Science

Citations listed below are presented in a standardized, modified format for display purposes only. They do not necessarily reflect the preferred style and conventions of the faculty member or discipline.

Rasmussen, Nicholas R., Rizk, Rodrigue, Wang, Longwei, Singh, Arun, and KC, Santosh. Channel-Selected Stratified Nested Cross-Validation for Clinically Relevant EEG-Based Parkinson’s Disease Detection. In 2026 IEEE Conference on Artificial Intelligence (CAI), 91-97. IEEE, 2026. (DOI: 10.1109/cai68641.2026.11536517)
Zhou, Zihan, Zhou, Yang, Yu, Jianghai , Lyu, Lingjuan, and Wang, Longwei. Structured Multi-step Jailbreaking under a Hamiltonian Generative Formulation. The 43rd International Conference on Machine Learning (ICML), 2026.
Wall, Casey, Wang, Longwei, Rizk, Rod, and Santosh, KC. Winsor-CAM: Human-tunable visual explanations from deep networks via layer-wise Winsorization. IEEE Transactions on Pattern Analysis and Machine Intelligence , 2026.
Zhang, Chaowei, Luo, Xiansheng, Qiang, Jipeng, and Wang, Longwei. Acting Flatterers via LLMs Sycophancy: Combating Clickbait with LLMs Opposing-Stance Reasoning. ACM Web Conference (WWW) , 2026.
Wang, Longwei, Uddin, Ifrat , and Santosh, KC. Expert-Guided Explainable Few-Shot Learning with Active Sample Selection for Medical Image Analysis. IEEE Journal of Biomedical and Health Informatics, 2026.
Peng, Sitian, Wang, Rui, and Wang, Longwei. MBD-Planner: A Real-Time Obstacle Avoidance Framework for UAVs via Feature-Domain Motion Blur Decoupling. IEEE Robotics and Automation Letters, 2026.
Wang, Longwei, Zhang, Chaowei, Uddin, Ifrat Ikhtear, Santosh, KC, Qin, Xiao, and Zhou, Yang. Bridging Symmetry and Robustness: On the Role of Equivariance in Enhancing Adversarial Robustness. Advances in Neural Information Processing Systems (NeurIPS), 2025.
Wang, Longwei, Nayyem, M., Rakin, A., Santosh, KC, Zhang, Chaowei, and Zhou, Yang. Explainability-Guided Defense: Attribution-Aware Model Refinement Against Adversarial Data Attacks. IEEE International Conference on Data Mining (ICDM), 2025.
KC, Santosh, Rizk, Rodrigue, and Wang, Longwei. Toward Carbon-Neutral Human AI: Rethinking Data, Computation, and Learning Paradigms for Sustainable Intelligence. In 2025 IEEE 7th International Conference on Cognitive Machine Intelligence (CogMI), 440-448. IEEE, 2025. (DOI: 10.1109/cogmi67134.2025.00058)
Wang, Longwei, Uddin, Ifrat I., Qin, Xiao, Zhou, Yang, and Santosh, KC. Explainability-Driven Defense: Grad-CAM-Guided Model Refinement Against Adversarial Threats, 49-57. Vol. 6, Iss. 1 Proceedings of the AAAI Symposium Series, 2025.
Ranabhat, Robin N., Wang, Longwei, Qin, Xiao, Zhou, Yang, and Santosh, KC. Multi-Scale Unrectified Push-Pull with Channel Attention for Enhanced Corruption Robustness, 34-41. Vol. 6, Iss. 1 Proceedings of the AAAI Symposium Series, 2025.
Ranabhat, Robin , Wang, Longwei, Patel, Amit, and Santosh, KC. Promoting Shape Bias in CNNs: Frequency-Based and Contrastive Regularization for Corruption Robustness. The 5th International Conference on Intelligent Systems and Pattern Recognition (Best Paper Award), 2025.
Uddin, Ifrat , Wang, Longwei, and Santosh, KC. Expert-guided explainable few-shot learning for medical image diagnosis, 95--104. MICCAI Workshop on Data Engineering in Medical Imaging, 2025.

Specialty

Contact Us