Bangli Liu

East China University of Science and Technology, University of Portsmouth

Papers

Total Citations

H-Index

About

Bangli Liu is a researcher whose work lies at the intersection of computer vision and human behavior analysis, with a particular focus on understanding and interpreting human attention and activity. Her most notable contribution is in the domain of visual focus of attention (VFOA) estimation, where she developed an improved method for estimating gaze directionality using eye center localization. This work, published in 2015 and garnering 27 citations, addresses a fundamental challenge in human-robot interaction systems by extracting reliable cues from eye movements. Liu’s approach to VFOA estimation has provided a practical foundation for systems that need to understand where a person is looking, enhancing the naturalness of human-machine communication. More recently, she has extended her research into the dynamic field of online activity recognition, proposing a multi-stage adaptive regression framework in 2019. This work demonstrates her commitment to real-time, adaptive systems that can learn and respond to human actions as they unfold. Through her research, Liu has made significant strides in bridging the gap between raw visual data and high-level behavioral understanding, contributing valuable tools for interactive and assistive technologies.