Graph R-CNN for Scene Graph Generation
arXiv:1808.00191
Abstract
We propose a novel scene graph generation model called Graph R-CNN, that is both effective and efficient at detecting objects and their relations in images. Our model contains a Relation Proposal Network (RePN) that efficiently deals with the quadratic number of potential relations between objects in an image. We also propose an attentional Graph Convolutional Network (aGCN) that effectively captures contextual information between objects and relations. Finally, we introduce a new evaluation metric that is more holistic and realistic than existing metrics. We report state-of-the-art performance on scene graph generation as evaluated using both existing and our proposed metrics.
arXiv comment16 pages, ECCV 2018 camera ready
References in corpus (6)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Pixels to Graphs by Associative Embedding
- Visual Relationship Detection with Language Priors
- Scene Graph Generation by Iterative Message Passing
- Detecting Visual Relationships with Deep Relational Networks
- Weakly-supervised learning of visual relations
Cited by in corpus (5)
- Action Genome: Actions as Composition of Spatio-temporal Scene Graphs
- Learning by Abstraction: The Neural State Machine
- Large-Scale Visual Relationship Understanding
- Learning Predicates as Functions to Enable Few-shot Scene Graph Prediction
- On Class Imbalance and Background Filtering in Visual Relationship Detection