Thomas Jantos
Papers
5
Total Citations
45
H-Index
3
About
Thomas Jantos is a leading researcher at the intersection of computer vision, robotics, and autonomous navigation, with a primary focus on 6-Degrees of Freedom (6-DoF) object pose estimation and AI-driven relative state estimation. His most impactful contribution is the **PoET (Pose Estimation Transformer)** framework, which tackles the challenging problem of single-view, multi-object 6D pose estimation for robotic grasping and localization, earning 28 citations. Jantos has made significant strides in automating data annotation for 6-DoF navigation algorithms, a critical bottleneck for developing robust AI systems for unmanned aircraft and autonomous driving. His recent work, **AIVIO**, introduces a closed-loop, object-relative navigation system for UAVs that integrates AI-aided visual-inertial odometry, enabling precise infrastructure inspection without GPS. Demonstrating versatility, Jantos also explores reinforcement learning for robotic control, as seen in his **CaRoSaC** framework for cable-driven parallel robots. With a growing citation record and a clear trajectory from foundational pose estimation to applied, self-calibrating navigation systems, Jantos is advancing the frontier of autonomous robotics in complex, real-world environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5