Hierarchical Representations and Explicit Memory: Learning Effective Navigation Policies on 3D Scene Graphs using Graph Neural Networks

Representations are crucial for a robot to learn effective navigation policies. Recent work has shown that mid-level perceptual abstractions, such as depth estimates or 2D semantic segmentation, lead to more effective policies when provided as observations in place of raw sensor data (e.g., RGB imag...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	arXiv.org 2022-05
Hauptverfasser:	Ravichandran, Zachary, Peng, Lisa, Hughes, Nathan, Griffith, J Daniel, Carlone, Luca
Format:	Artikel
Sprache:	eng
Schlagworte:	Ablation Color imagery Computer architecture Graph neural networks Graphical representations Graphs Image segmentation Learning Navigation Neural networks Occupancy Policies Robots Semantics Topology
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!