Visual Semantic Reasoning for Image-Text Matching
arXiv:1909.02701
Abstract
Image-text matching has been a hot research topic bridging the vision and language areas. It remains challenging because the current representation of image usually lacks global semantic concepts as in its corresponding text caption. To address this issue, we propose a simple and interpretable reasoning model to generate visual representation that captures key objects and semantic concepts of a scene. Specifically, we first build up connections between image regions and perform reasoning with Graph Convolutional Networks to generate features with semantic relationships. Then, we propose to use the gate and memory mechanism to perform global semantic reasoning on these relationship-enhanced features, select the discriminative information and gradually generate the representation for the whole scene. Experiments validate that our method achieves a new state-of-the-art for the image-text matching on MS-COCO and Flickr30K datasets. It outperforms the current best method by 6.8% relatively for image retrieval and 4.8% relatively for caption retrieval on MS-COCO (Recall@1 using 1K test set). On Flickr30K, our model improves image retrieval by 12.6% relatively and caption retrieval by 5.8% relatively (Recall@1). Our code is available at https://github.com/KunpengLi1994/VSRN.
arXiv commentAccepted to ICCV 2019 (Oral)
References in corpus (7)
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
- Unifying Visual-Semantic Embeddings with Multimodal Neural Language Models
- A simple neural network module for relational reasoning
- Visual Semantic Navigation using Scene Priors
- Iterative Visual Reasoning Beyond Convolutions
- Multimodal Style Transfer via Graph Cuts
Cited by in corpus (4)
- TransRefer3D: Entity-and-Relation Aware Transformer for Fine-Grained 3D Visual Grounding
- Cops-Ref: A new Dataset and Task on Compositional Referring Expression Comprehension
- Constructing Multi-Modal Dialogue Dataset by Replacing Text with Semantically Relevant Images
- Exploiting Visual Semantic Reasoning for Video-Text Retrieval