Enhancing Scene Graph Generation with Hierarchical Relationships and Commonsense Knowledge
This work addresses the problem of generating more accurate and sensible scene graphs for computer vision applications, representing an incremental enhancement to existing methods.
The paper tackles scene graph generation by incorporating hierarchical relationships and commonsense knowledge, resulting in significant improvements with extensive reasonable predictions beyond dataset annotations on Visual Genome and OpenImage V6 datasets.
This work introduces an enhanced approach to generating scene graphs by incorporating both a relationship hierarchy and commonsense knowledge. Specifically, we begin by proposing a hierarchical relation head that exploits an informative hierarchical structure. It jointly predicts the relation super-category between object pairs in an image, along with detailed relations under each super-category. Following this, we implement a robust commonsense validation pipeline that harnesses foundation models to critique the results from the scene graph prediction system, removing nonsensical predicates even with a small language-only model. Extensive experiments on Visual Genome and OpenImage V6 datasets demonstrate that the proposed modules can be seamlessly integrated as plug-and-play enhancements to existing scene graph generation algorithms. The results show significant improvements with an extensive set of reasonable predictions beyond dataset annotations. Codes are available at https://github.com/bowen-upenn/scene_graph_commonsense.