What Makes a Photograph Memorable?

Abstract

Abstract-When glancing at a magazine, or browsing the Internet, we are continuously exposed to photographs. Despite this overflow of visual information, humans are extremely good at remembering thousands of pictures along with some of their visual details. But not all images are equal in memory. Some stitch to our minds while others are quickly forgotten. In this paper we focus on the problem of predicting how memorable an image will be. We show that memorability is an intrinsic and stable property of an image that is shared across different viewers, and remains stable across delays. We introduce a database for which we have measured the probability that each picture will be recognized after a single view. We analyze a collection of image features, labels, and attributes that contribute to making an image memorable, and we train a predictor based on global image descriptors. We find that predicting image memorability is a task that can be addressed with current computer vision techniques. While making memorable images is a challenging task in visualization, photography, and education, this work is a first attempt to quantify this useful property of images.

Understanding and Predicting Image Memorability at a Large Scale

“…Interestingly, we also tend to remember and forget the same pictures and faces as each other [1,13]. This suggests that despite different personal experiences, people naturally encode and discard the same types of information.…”

“…Here, we introduce an optimized protocol of the memory game introduced by [13] to collect human memory scores. In this game, images are presented successively, and some are repeated.…”

“…The basic idea of our novel procedure is to allow the second occurrence of an image to occur at variable time intervals. This procedure is based on the finding that the memorability ranks of images are time-independent [13]. We propose an algorithm to account for this varied time interval allowing us to obtain high consistency with the existing benchmark [13].…”

“…This procedure is based on the finding that the memorability ranks of images are time-independent [13]. We propose an algorithm to account for this varied time interval allowing us to obtain high consistency with the existing benchmark [13]. Furthermore, using this new experimental setting, we build a novel massive memorability dataset, with scores on 60, 000 images (∼ 27 times the previous largest benchmark), while keeping a low cost.…”

“…Combining the versatility of many benchmarks and a novel experimental method for efficiently collecting human memory scores (about one-tenth the cost of [13]), we introduce the LaMem dataset, containing 60,000 images with memorability scores from human observers (about 27 times larger than the previous dataset [13]). …”