Don't only Feel Read: Using Scene text to understand advertisements
This work addresses the challenge of automated advertisement classification for marketing or content analysis, but it is incremental as it uses off-the-shelf components without introducing new methods.
The paper tackles the problem of classifying advertisement images by incorporating textual cues from embedded text alongside visual features, demonstrating that textual content improves classification performance.
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.