1University of Illinois at Urbana-Champaign
2Georgia Institute of Technology
This paper proposes a method for learning joint embeddings of images and text using a two-branch neural network with multiple layers of linear projections followed by nonlinearities. The network is trained using a large-margin objective that combines cross-view ranking constraints with within-view neighborhood structure preservation constraints inspired by metric learning literature. Extensive experiments show that our approach gains significant improvements in accuracy for image-to-text and text-to-image retrieval. Our method achieves new state-of-the-art results on the Flickr30K and MSCOCO image-sentence datasets and shows promise on the new task of phrase localization on the Flickr30K Entities dataset.
Deep Structure-Preserving Image-Text Embeddings.
We released the matlab version code of our two branch deep embedding method for only academic research use. For instructions of this code, please see Readme.txt in the following folder.
Any questions, please feel free to contact Liwei Wang, firstname.lastname@example.org