Amir Geva

University of Fukui, University of California, Santa Barbara, Technion – Israel Institute of Technology

Papers

Total Citations

H-Index

About

Amir Geva is a researcher in robotics and computer vision, specializing in vision-based localization for autonomous vehicles. His work focuses on solving the critical problem of indoor positioning for flying and ground robots, where traditional methods like GPS or odometry are unavailable. Geva’s major contributions center on leveraging prior knowledge—such as digital terrain models and building floorplans—to enhance camera pose estimation. His most-cited paper (2015, 10 citations) advances Bundle Adjustment for simultaneous localization and mapping (SLAM), addressing the unique challenges of aerial robots by incorporating terrain constraints. He further developed a vision-based indoor positioning system (2018, 6 citations) that uses a monocular camera and a floorplan to determine a vehicle’s location without requiring dense 3D maps or prior environment exploration. His 2019 follow-up (6 citations) extended this to global positioning from a known start, demonstrating robust performance even when scene geometry or lighting changes. Geva’s work is notable for its practical, minimal-infrastructure approach, making it highly relevant for real-world deployment in GPS-denied environments. His research bridges the gap between theoretical SLAM advances and deployable robotic systems, with applications in logistics, inspection, and search-and-rescue.