Hai-Feng Pi

Papers

Total Citations

H-Index

About

Hai-Feng Pi is a researcher whose work lies at the intersection of computer vision and human-computer interaction, with a particular focus on gaze estimation—a critical technology for understanding human attention in applications ranging from virtual reality to accessibility tools. His most notable contribution, the 2022 paper "Style transformed synthetic images for real world gaze estimation by using residual neural network with embedded personal identities," has garnered 9 citations, reflecting its innovative approach to bridging the gap between synthetic training data and real-world performance. Pi’s key insight involves using style transformation techniques to make synthetic images more realistic, combined with a residual neural network that embeds personal identity information, thereby improving the accuracy and generalization of gaze estimation models across diverse users. This work addresses a fundamental challenge in the field: the domain shift between controlled training datasets and unpredictable real-world environments. By tackling this issue, Pi has advanced the practical deployment of gaze-tracking systems, making them more robust and user-adaptive. His research is particularly valuable for students and engineers seeking to understand how deep learning can be leveraged to solve real-world perception problems, and it positions him as a thoughtful contributor to the ongoing evolution of human-centered AI.