English

Don't only Feel Read: Using Scene text to understand advertisements

Computer Vision and Pattern Recognition 2019-11-14 v3

Abstract

We propose a framework for automated classification of Advertisement Images, using not just Visual features but also Textual cues extracted from embedded text. Our approach takes inspiration from the assumption that Ad images contain meaningful textual content, that can provide discriminative semantic interpretetion, and can thus aid in classifcation tasks. To this end, we develop a framework using off-the-shelf components, and demonstrate the effectiveness of Textual cues in semantic Classfication tasks.

Keywords

Cite

@article{arxiv.1806.08279,
  title  = {Don't only Feel Read: Using Scene text to understand advertisements},
  author = {Arka Ujjal Dey and Suman K. Ghosh and Ernest Valveny},
  journal= {arXiv preprint arXiv:1806.08279},
  year   = {2019}
}

Comments

Accepted in CVPR 2018 Workshop: Towards Automatic Understanding of Visual Advertisements (ADS)

R2 v1 2026-06-23T02:37:24.904Z