In-Image Advertising

Abstract

The daunting volumes of images on the Web have become one of the primary sources for online advertising. This work presents a contextual in-image advertising strategy driven by images, which automatically associates relevant ads with an image and seamlessly inserts the ads in the nonintrusive areas within each individual image. In in-image advertising platform, the ads are selected based on not only textual relevance but also visual similarity. The ad insertion positions are detected based on image salience, as well as face detection, to minimize intrusiveness to the user. In addition to general in-image advertising, we also provide a special game-like in-image advertising model dedicated to image on the basis of gaming form, called GameSense, which supports creating a game from an online image and associates relevant ads within the game. We evaluate in-image advertising model on a large-scale real-world images, and demonstrate the effectiveness of in-image advertising platform.