Trustworthy AI Systems Lab (TAS Lab)

The Trustworthy AI Systems Lab (TAS Lab) studies robust machine learning, trustworthy perception, secure intelligent systems, and dependable AI deployment.

We develop methods and systems that help AI behave reliably under distribution shift, adversarial pressure, and real-world deployment constraints.

Publications

A concealed poisoning attack to reduce deep neural networks' robustness against adversarial samples
Information Sciences, 758-773 2022
Junhao Zheng, Patrick Chan, Huiyang Chi, Zhimin He
Trustworthy AI Poisoning Attack Adversarial Learning Robust Machine Learning
Progressive editing with stacked Generative Adversarial Network for multiple facial attribute editing Cite
Computer Vision and Image Understanding 2022
Patrick Chan, Xiaotian Wang, Zhe Lin, Daniel S. Yeung
Computer Vision GAN Facial Attribute Editing
Unsupervised Domain Adaptation for Gesture Identification against Electrode Shift PDF Cite Code
IEEE Transactions on Human-Machine Systems 2022
Patrick Chan, Qiuxia Li, Yingfeng Fang, Linyi Xu, Kairu Li, Honghai Liu, Daniel S. Yeung
Domain Adaptation Gesture Recognition Human-Computer Interaction
Class-Specific Affinity based Weakly Supervised Semantic Segmentation with Neutral Region Exploration
IJCNN 2021 2021
Keke Chen, Patrick Chan, Tianyi Xiang, Natasha Kees, Daniel S. Yeung
Computer Vision Semantic Segmentation Weakly Supervised Learning Metric Learning
Distribution-based Adversarial Filter Feature Selection against Evasion Attack
IJCNN 2021 2021
Patrick Chan, YuanChao Liang, Fei Zhang, Daniel S. Yeung
Trustworthy AI Adversarial Learning Feature Selection Evasion Attack
See all publications