Neural Motifs: Scene Graph Parsing with Global Context

TL;DR

This paper explores the structure of scene graphs in visual scenes, revealing recurring motifs and introducing a new architecture that leverages these motifs to improve scene graph parsing accuracy.

Contribution

The paper provides new insights into motif patterns in scene graphs and proposes Stacked Motif Networks to better capture higher order motifs, improving parsing performance.

Findings

Object labels strongly predict relation labels.

Over 50% of scene graphs contain motifs with multiple relations.

The proposed architecture outperforms previous methods by up to 7.1%.

Abstract

We investigate the problem of producing structured graph representations of visual scenes. Our work analyzes the role of motifs: regularly appearing substructures in scene graphs. We present new quantitative insights on such repeated structures in the Visual Genome dataset. Our analysis shows that object labels are highly predictive of relation labels but not vice-versa. We also find that there are recurring patterns even in larger subgraphs: more than 50% of graphs contain motifs involving at least two relations. Our analysis motivates a new baseline: given object detections, predict the most frequent relation between object pairs with the given labels, as seen in the training set. This baseline improves on the previous state-of-the-art by an average of 3.6% relative improvement across evaluation settings. We then introduce Stacked Motif Networks, a new architecture designed to capture higher order motifs in scene graphs that further improves over our strong baseline by an average 7.1% relative gain. Our code is available at github.com/rowanz/neural-motifs.