Constructing a Visual Relationship Authenticity Dataset
This dataset addresses a gap in visual relationship detection for scene understanding and grounded natural language processing, though it is incremental as it builds on existing datasets.
The authors tackled the problem of distinguishing false visual relationships from true ones in images by constructing a visual relationship authenticity dataset, which includes both true and false annotations based on the Flickr30k entities dataset and is publicly available.
A visual relationship denotes a relationship between two objects in an image, which can be represented as a triplet of (subject; predicate; object). Visual relationship detection is crucial for scene understanding in images. Existing visual relationship detection datasets only contain true relationships that correctly describe the content in an image. However, distinguishing false visual relationships from true ones is also crucial for image understanding and grounded natural language processing. In this paper, we construct a visual relationship authenticity dataset, where both true and false relationships among all objects appeared in the captions in the Flickr30k entities image caption dataset are annotated. The dataset is available at https://github.com/codecreator2053/VR_ClassifiedDataset. We hope that this dataset can promote the study on both vision and language understanding.