Max Robotics — Global Service Robot Directory + US Certification

About

Dongil Han is a leading researcher in computer vision and human-robot interaction, with key contributions spanning stereo vision, real-time face detection, and affective computing. His early work on stereo matching introduced a novel method for wide disparity range detection, laying groundwork for depth perception in robotic systems. Han is perhaps best known for pioneering high-performance, real-time face detection hardware architectures. His 2010 paper, with 8 citations, proposed an FPGA-based engine using Modified Census Transform (MCT) that operates at over 60 frames per second, robust to illumination changes and rotated faces—critical for robot vision. This work, along with his 2011 follow-up, established a foundation for embedded vision systems. More recently, Han has advanced multimodal emotion recognition for human-robot interaction, introducing a hierarchical attention approach (2021, 7 citations) that addresses the lack of contextual understanding in emotional perception. His research on projection-based region merging for disparity map segmentation (2007, 3 citations) further demonstrates his impact on household robotics. With cumulative citations exceeding 38 across his most-cited works, Han’s contributions continue to influence real-time, robust vision systems that enable more natural and reliable human-robot collaboration.