INGRESS: Interactive visual grounding of referring expressions

This article presents INGRESS, a robot system that follows human natural language instructions to pick and place everyday objects. The key question here is to ground referring expressions: understand expressions about objects and their relationships from image and natural language inputs. INGRESS al...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	The International journal of robotics research 2020-03, Vol.39 (2-3), p.217-232
Hauptverfasser:	Shridhar, Mohit, Mittal, Dixant, Hsu, David
Format:	Artikel
Sprache:	eng
Schlagworte:	Natural language Neural networks Questions Robots Source code
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!