Max Robotics — Global Service Robot Directory + US Certification

About

Andreas Doumanoglou is a robotics and computer vision researcher whose work sits at the intersection of object pose estimation, robotic manipulation, and machine learning. He is perhaps best known for his influential contributions to 6D object pose estimation, particularly in complex, cluttered scenes — a problem critical to advancing robotics and augmented reality. His 2016 paper "Recovering 6D Object Pose and Predicting Next-Best-View in the Crowd" has accumulated over 230 citations, establishing him as a leading voice in handling severe occlusions and multi-instance detection challenges. Equally significant is Doumanoglou's pioneering research on autonomous robotic clothing manipulation. His work on recognizing, unfolding, and folding garments using dual-armed robots — leveraging random decision forests and probabilistic planning — tackled one of robotics' most notoriously difficult manipulation problems due to the deformable, unpredictable nature of fabric. These papers have collectively garnered over 300 citations, reflecting their broad influence across the robotics community. His development of Active Random Forests further demonstrates a commitment to efficient, intelligent sensing strategies. Across his body of work, Doumanoglou consistently bridges theoretical machine learning with real-world robotic applications, making his research highly relevant for students and engineers working in autonomous systems and embodied AI.