SEM-GAT: explainable semantic pose estimation using learned graph attention

This paper proposes a Graph Neural Network (GNN)-based method for exploiting semantics and local geometry to guide the identification of reliable pointcloud registration candidates. Semantic and morphological features of the environment serve as key reference points for registration, enabling accura...

תיאור מלא

מידע ביבליוגרפי
Main Authors: Panagiotaki, E, De Martini, D, Pramatarov, G, Gadd, M, Kunze, L
פורמט: Conference item
שפה:English
יצא לאור: IEEE 2024