English

Multitask Text-to-Visual Embedding with Titles and Clickthrough Data

Computer Vision and Pattern Recognition 2019-06-03 v1 Information Retrieval

Abstract

Text-visual (or called semantic-visual) embedding is a central problem in vision-language research. It typically involves mapping of an image and a text description to a common feature space through a CNN image encoder and a RNN language encoder. In this paper, we propose a new method for learning text-visual embedding using both image titles and click-through data from an image search engine. We also propose a new triplet loss function by modeling positive awareness of the embedding, and introduce a novel mini-batch-based hard negative sampling approach for better data efficiency in the learning process. Experimental results show that our proposed method outperforms existing methods, and is also effective for real-world text-to-visual retrieval.

Keywords

Cite

@article{arxiv.1905.13339,
  title  = {Multitask Text-to-Visual Embedding with Titles and Clickthrough Data},
  author = {Pranav Aggarwal and Zhe Lin and Baldo Faieta and Saeid Motiian},
  journal= {arXiv preprint arXiv:1905.13339},
  year   = {2019}
}

Comments

4 pages. Language and Vision Workshop, in conjunction with CVPR 2019

R2 v1 2026-06-23T09:34:14.061Z