In Defense of Scene Graph Generation for Human-Robot Open-Ended Interaction in Service Robotics - Archive ouverte HAL
Communication Dans Un Congrès Année : 2024

In Defense of Scene Graph Generation for Human-Robot Open-Ended Interaction in Service Robotics

Résumé

Compositional relations represent a good source of information in the task of scene understanding. However, current approaches in domestic service robotics only scratch the surface of the benefits of compositional relations by leveraging only their spatial component. In this position paper, we propose a new perspective on the use of compositional relations as a means to extract meaning from context in open-ended interactions. We especially design a multi-layer representation based on scene graphs that encapsulates four different dimensions of knowledge. To exploit this new representation, we introduce a new large-scale dataset for indoor service robots with high-quality scene graph annotations. We then argue for the opportunities of using this representation to easily extract a wide range of fine-grained information about human interaction with context (All data and code are available at https://github.com/Maelic/IndoorVG).
Fichier non déposé

Dates et versions

hal-04585002 , version 1 (23-05-2024)

Identifiants

Citer

Maëlic Neau, Paulo Santos, Anne-Gwenn Bosser, Cédric Buche. In Defense of Scene Graph Generation for Human-Robot Open-Ended Interaction in Service Robotics. Robot World Cup 2023, Jul 2023, Bordeaux, France. pp.299-310, ⟨10.1007/978-3-031-55015-7_25⟩. ⟨hal-04585002⟩
14 Consultations
0 Téléchargements

Altmetric

Partager

More