Max Robotics — Global Service Robot Directory + US Certification

About

Sergey Linok is a robotics and computer vision researcher whose work sits at the intersection of autonomous navigation, scene understanding, and intelligent robotic control. His research spans several interconnected areas, including 3D object grounding, self-supervised depth estimation, and mobile manipulation in human-centered environments. Linok's most impactful contribution is his work on open-vocabulary object grounding using 3D scene graphs, which addresses a critical limitation of existing CLIP-based methods — their inability to resolve ambiguous natural language descriptions that require relational and contextual scene understanding. This work has garnered 11 citations and represents a meaningful advance for autonomous agents operating in complex real-world settings. His analysis of neural network receptive fields for monocular depth and ego-motion estimation (6 citations) contributes valuable insights to robot perception systems relying solely on single-camera inputs. Through his involvement in the STRL Robotics project (5 citations), Linok has also demonstrated expertise in integrating localization, mapping, and motion planning within unified robotic control architectures. His work on door-opening strategies for mobile manipulators further highlights his practical focus on constrained, real-world robotic challenges. Across his portfolio, Linok consistently bridges theoretical computer vision with applied robotics, making his research particularly relevant for developers of next-generation autonomous systems.