Cem Karaoguz

Bielefeld University, Institut national de recherche en sciences et technologies du numérique, Honda (Germany)

Papers

Total Citations

H-Index

About

Cem Karaoguz’s research lies at the intersection of robotics, active vision, and machine learning, with a focus on enabling autonomous systems to perceive and act more intelligently in complex, noisy environments. His major contributions center on how robots can learn to integrate multiple sensory cues—such as audio and visual depth information—through reward-based learning, as demonstrated in his most-cited 2011 paper (6 citations). This work addresses a fundamental challenge in robotics: improving precision by optimally combining imprecise data sources. Karaoguz also pioneered incremental learning techniques to bootstrap object classifier models using synthetic data from virtual reality tools (2016, 4 citations), significantly reducing the costly data collection burden for real-world applications. Additionally, his 2011 study on optimizing gaze movement for multitasking (2 citations) advanced active vision approaches, enabling robots to efficiently allocate visual attention during tasks like grasping or navigation. By tackling core issues in sensor fusion, data efficiency, and attentional control, Karaoguz’s work has laid important groundwork for more adaptive, resource-aware robotic systems, making him a notable contributor to the fields of cognitive robotics and autonomous perception.