< Back to previous page

Project

The Visual Narrative: Conveying Meaning through Visual Content (R-11234)

Images are an integral part of human communication. From cave drawings to digital photography, they have been largely used to inform, persuade, and entertain. However, while some images can tell a compelling story to the viewer, others are doomed to be left unnoticed. The ability of the image to transfer a message has long been a topic of research in psychology and art. Recently, visual data have also become a focus of computer vision scholars. They automatically analyzed images to discover what makes them aesthetically pleasing (Datta et al., 2006), interesting (Dhar et al., 2011), and memorable (Khosla et al., 2012). However, to the best of our knowledge, no prior research used automatic image analysis to reveal how visual content conveys meaning. The present research attempts to fill this void. We expect to contribute to the existing studies on visual data analysis in three main domains. Our first foundational study is expected to open a black box of visual cognition by revealing what is in an image that affects viewers' attitude and behavior. In the second project, we aim to explore how to efficiently encode meaning in an image to facilitate message transfer. Finally, using automatic image analysis in the third study, we will complete the research by investigating how to best decode the meaning from an image to better track and predict the needs, wants, and intentions of viewers.
Date:1 Nov 2020 →  31 Oct 2022
Keywords:deep learning, machine learning, social media
Disciplines:Marketing not elsewhere classified