Unbiased Heterogeneous Scene Graph Generation with Relation-aware Message Passing Neural Network

TL;DR

This paper introduces HetSGG, a novel framework for scene graph generation that uses relation-aware message passing neural networks to better model context and improve performance, especially on rare predicate classes.

Contribution

The paper proposes a relation-aware message passing neural network (RMP) layer for heterogeneous scene graph generation, addressing the limitations of previous homogeneous graph assumptions.

Findings

HetSGG outperforms state-of-the-art methods.

Significant improvement on tail predicate classes.

Effective modeling of relation-dependent context.

Abstract

Recent scene graph generation (SGG) frameworks have focused on learning complex relationships among multiple objects in an image. Thanks to the nature of the message passing neural network (MPNN) that models high-order interactions between objects and their neighboring objects, they are dominant representation learning modules for SGG. However, existing MPNN-based frameworks assume the scene graph as a homogeneous graph, which restricts the context-awareness of visual relations between objects. That is, they overlook the fact that the relations tend to be highly dependent on the objects with which the relations are associated. In this paper, we propose an unbiased heterogeneous scene graph generation (HetSGG) framework that captures relation-aware context using message passing neural networks. We devise a novel message passing layer, called relation-aware message passing neural network (RMP), that aggregates the contextual information of an image considering the predicate type between objects. Our extensive evaluations demonstrate that HetSGG outperforms state-of-the-art methods, especially outperforming on tail predicate classes.