Ken Ito

Papers

Total Citations

H-Index

About

Ken Ito’s research lies at the intersection of computer vision and robotics, with a primary focus on visual tracking systems that enable machines to perceive and follow objects in dynamic environments. His major contributions center on developing view-based tracking methods that robustly handle appearance changes caused by 3D motion and occlusions. Ito pioneered the use of affine transformed templates to model how an object’s appearance shifts under different viewpoints, allowing tracking systems to maintain accuracy even when the target rotates, scales, or becomes partially hidden. His 2002 paper on robust view-based visual tracking with occlusion detection, which has garnered 14 citations, introduced critical techniques for identifying and recovering from occluded states—a persistent challenge in real-world robotic tasks like object manipulation and mobile robot navigation. Earlier, his 2001 work on dynamic transitions in groups of affine transformed templates (7 citations) laid the groundwork for adaptive template selection, enabling smoother tracking across abrupt appearance changes. Ito’s research has directly influenced the development of more resilient vision systems for autonomous robots, demonstrating how principled geometric modeling can overcome the limitations of fixed-template approaches. His work remains a foundational reference for researchers tackling visual tracking in unstructured, unpredictable environments.