Understanding Image Memorability through Localized Stimuli

Document Type : Research Paper

Authors

1 Electrical & Computer Engineering Department, Semnan University, Semnan, Iran

2 Faculty of Electrical and Computer Engineering (ECE), Semnan University, Semnan, Iran.

10.22075/mseee.2024.33159.1143

Abstract

There are different images whenever people read a magazine, browse the Internet, or use social media. While humans can recall thousands of images in everyday life along with their visual details, not all images in their memory are the same, even though they have access to an abundance of visual information daily. Other researchers have examined how image features affect perception. In our study, we explore how local images affect image memorability. First, we introduce new datasets with local images. Secondly, we design a visual memory game that utilizes a human agent to quantify image memorability. Finally, we predict memorability scores with ResNet 50 and ResNet 101. The result was that local images impacted image memorability.

Keywords

Main Subjects



Articles in Press, Accepted Manuscript
Available Online from 01 July 2024
  • Receive Date: 02 February 2024
  • Revise Date: 27 March 2024
  • Accept Date: 01 July 2024