Mariane Giambastiani

Universidade Federal do Rio Grande do Sul

Papers

Total Citations

H-Index

About

Mariane Giambastiani is a roboticist specializing in autonomous navigation, simultaneous localization and mapping (SLAM), and topological mapping. Her research focuses on enabling robots to build accurate maps of unknown environments in real time, a critical capability for autonomous exploration. Her most influential work, "Ouroboros: Using potential field in unexplored regions to close loops" (2015, 15 citations), introduces an innovative loop-closing strategy for integrated exploration. By leveraging potential fields in unexplored areas, this approach allows robots to efficiently detect and close loops in their trajectories, significantly improving map consistency without requiring prior environmental knowledge. Giambastiani also contributed to place recognition for topological SLAM with her paper "Using n-grams of spatial densities to construct maps" (2015), which explores how spatial density patterns can be used to robustly associate vertices in topological representations. Her work addresses fundamental challenges in autonomous robotics, particularly the front-end perception problems that underpin reliable SLAM systems. While her citation counts reflect a focused research output, her contributions to loop-closing and topological mapping represent meaningful advances in how robots perceive and navigate unknown spaces.