Learning Visual Relation Priors for Image-Text Matching and Image Captioning with Neural Scene Graph Generators

Grounding language to visual relations is critical to various language-and-vision applications. In this work, we
tackle two fundamental language-and-vision tasks: image-text matching and image captioning, and demonstrate that
neural scene graph generators can learn effective visual relation features to facilitate grounding language to visual
relations and subsequently improve the two end applications. By combining relation features with the state-of-the-art models, our experiments show significant improvement on the standard Flickr30K and MSCOCO benchmarks. Our
experimental results and analysis show that relation features improve downstream models’ capability of capturing visual relations in end vision-and-language applications. We also demonstrate the importance of learning scene graph generators with visually relevant relations to the effectiveness of relation features.