DDS: Decoupled Dynamic Scene-Graph Generation Network
This addresses a bottleneck in scene understanding for computer vision applications, though it appears incremental as it builds on existing scene-graph generation methods.
The paper tackles the problem of scene-graph generation's poor performance on novel object-relationship triplets by proposing DDS, a decoupled dynamic network that disentangles object and relationship features, resulting in significant performance gains on three datasets, particularly for unseen triplets.
Scene-graph generation involves creating a structural representation of the relationships between objects in a scene by predicting subject-object-relation triplets from input data. Existing methods show poor performance in detecting triplets outside of a predefined set, primarily due to their reliance on dependent feature learning. To address this issue, we propose DDS -- a decoupled dynamic scene-graph generation network -- that consists of two independent branches that can disentangle extracted features. The key innovation of the current paper is the decoupling of the features representing the relationships from those of the objects, which enables the detection of novel object-relationship combinations. The DDS model is evaluated on three datasets and outperforms previous methods by a significant margin, especially in detecting previously unseen triplets.