Our research focuses on assessing and predicting the perceptual quality of images and videos. By modeling human visual perception, we develop quality assessment methods that evaluate visual fidelity and naturalness, aligning automated evaluation with human judgments across diverse multimedia applications.

Our research focuses on developing intelligent video analytics algorithms for object detection, multi-object tracking, and human activity understanding. By analyzing spatial and temporal patterns, we detect abnormal events and assess crowd dynamics, supporting intelligent surveillance and safety monitoring across diverse real-world environments.

Our research focuses on assessing fairness and ethical implications in visual AI systems. We develop methods to evaluate algorithmic bias, privacy risks, and visual content authenticity, promoting responsible AI development through transparent assessment and human oversight.