Max Robotics — Global Service Robot Directory + US Certification

About

Longin Jan Latecki is a prominent computer scientist whose research spans computer vision, shape analysis, and autonomous robotics. His work bridges theoretical foundations in geometric shape representation with practical applications in self-driving vehicles and mobile robot navigation, making him a versatile and impactful figure in his field. Latecki's early contributions established influential frameworks for shape similarity and partial shape matching, with his 2004 work on optimal partial shape similarity and related investigations into visual parts and contour grouping laying important groundwork for object recognition. His robotics-focused research demonstrated particular depth, addressing multi-robot mapping, laser range data segmentation, and 3D planar structure extraction — collectively advancing how autonomous systems perceive and model their surroundings. His polygonal approximation and EM-based segmentation methods offered elegant, computationally efficient solutions to sensor data interpretation challenges faced by mobile robots. More recently, Latecki contributed to real-time semantic segmentation for autonomous driving through AGLNet, his most cited work with 129 citations, reflecting the field's growing demand for lightweight yet accurate deep learning architectures. His work on multiple target detection and tracking in cluttered indoor environments further demonstrates his commitment to practically deployable vision systems. Across more than two decades of research, Latecki has consistently advanced the intersection of geometric reasoning, machine learning, and robotic perception.