Hi there! My name is Wei Huang. I am currently a researcher at Purple Mountain Laboratories. I obtained my PhD in the Computer Science Department at the University of Liverpool and was supervised by Dr. Xiaowei Huang and Dr. Xingyu Zhao. Prior to study at University of Liverpool, I received my MSc degree at Imperial College London and BSc degree at Xiamen University.

Reseach Interest

My research interest remains in the field of AI safety. Machine learning models are suffering from the adversarial threats, like evasion attack, backdoor attack. I investigate into the attack and defence algorithms and develop the testing and verification methods for different machine learning models, including Ensemble Trees, CNNs, RNNs etc.

Recent News

• (06/2024) The project titled “Key Technological Research on the Endogenous Safety and Security of Deep Learning Models” has received funding under the 2024 Jiangsu Province Excellent Postdoctoral Program.
• (03/2024) Paper “Diversity Supporting Robustness: Enhancing Adversarial Robustness via Differentiated Ensemble Predictions” has been accepted to Computers & Security.
• (04/2024) Paper “Formal Verification of Robustness and Resilience of Learning-Enabled State Estimation Systems” has been accepted to Neurocomputing.
• (09/2023) Paper “Hierarchical Distribution-Aware Testing of Deep Learning” has been accepted to TOSEM.
• (07/2023) Paper “SAFARI: Versatile and Efficient Evaluations for Robustness of Interpretability” has been accepted to ICCV 2023.