Gauthier Lafruit

Papers

Total Citations

H-Index

About

Gauthier Lafruit is a leading researcher in real-time computer vision and immersive telepresence, with a career spanning high-performance stereo matching to cutting-edge virtual reality. His most influential work focuses on achieving both speed and accuracy in stereo vision, a critical challenge for autonomous systems and robotics. His seminal papers, "Real-time accurate stereo with bitwise fast voting on CUDA" (2009) and its scalable follow-up (2011), have collectively garnered nearly 100 citations. In these works, Lafruit pioneered a GPU-optimized, bitwise fast voting method that dramatically accelerates local stereo matching algorithms while preserving high accuracy through shape-adaptive cost aggregation. This breakthrough enabled real-time depth perception on parallel architectures, laying foundational groundwork for modern embedded vision systems. More recently, Lafruit has pushed the boundaries of tele-robotics, co-authoring a 2022 study on "Tele-Robotics VR with Holographic Vision in Immersive Video." This first-of-its-kind system integrates force feedback with holographic vision, allowing a user to remotely operate a robot arm while being virtually immersed in the scene—a significant leap toward seamless human-robot interaction. His work consistently bridges algorithmic innovation with practical, real-time deployment.