Max Robotics — Global Service Robot Directory + US Certification

About

Petra Bosilj is a leading researcher at the intersection of computer vision, robotics, and precision agriculture. Her work focuses on developing deep learning methods for semantic segmentation, object detection, and counting—enabling agricultural robots to distinguish crops from weeds, estimate fruit yields, and monitor biodiversity. Her most cited paper (161 citations) demonstrates transfer learning between crop types for semantic segmentation in precision agriculture, a foundational contribution that reduces chemical usage while improving crop health. She has also advanced robotic phenotyping by comparing deep regression and detection methods for counting fruit and grains (25 citations), and contributed to the emerging field of biodiversity monitoring with autonomous systems (13 citations). Her work on domain-generalised one-stage detection (4 citations) addresses the critical challenge of real-time vision under varying conditions, while her exploration of self-supervised learning for potato instance segmentation (2 citations) pushes data-efficient methods forward. Bosilj’s research is pivotal in making agricultural robotics more accurate, adaptable, and sustainable, with clear impact on both food production and environmental conservation.