Isaac Esteban

Papers

Total Citations

H-Index

About

Isaac Esteban is a roboticist whose work centers on the critical challenge of long-term visual odometry and large-scale mapping. His research addresses the fundamental problem of drift in integrated pose estimation, where small errors in relative motion measurements accumulate over time, causing trajectories to diverge. Esteban’s most influential contribution, detailed in his highly cited 2010 paper “Efficient trajectory bending with applications to loop closure” (20 citations), introduces a novel method for correcting these accumulated errors. By recognizing that a robot’s trajectory can be “bent” to satisfy loop closure constraints, his work provides an elegant and computationally efficient solution for maintaining global consistency in mapping. This approach is particularly vital for autonomous navigation in large, feature-sparse environments. Earlier, in his 2009 work “Mapping large environments with an omnivideo camera” (5 citations), Esteban explored the use of omnidirectional cameras and an Extended Information Filter to build maps from a trajectory-based framework. Collectively, his research provides foundational techniques for robust, long-duration robot autonomy, directly impacting fields from service robotics to autonomous exploration.