| Isola, P., Xiao, J., Torralba, A., & Oliva, A. (in press). Proceedings of the 24rd IEEE Conference on Computer Vision and Pattern Recognition. |
|
When glancing at a magazine, or a book, we are continuously being exposed to photographs. Despite of this overflow of visual information, humans are extremely good at remembering thousands of pictures along with their visual details ( Brady et al., PNAS 2008 ; Konkle et al., JEP:G 2010 , Konkle, et al., Psychological Science 2010 ). But not all images are created equal. Some images stitch to our minds, and we are able to recognize them even after longs periods of time. In this paper we focus on the problem of predicting how memorable an image will be. As opposed to other subjective image properties, image memorability has not been addressed before. Making memorable images is a challenging task in visualization and photography, and is generally presented as a vague concept hard to quantify. Surprisingly, there has been no previous attempt to systematically measure this property on image collections, and to apply computer vision techniques to extract memorability automatically.
Here we show that the memorability of a photograph is a stable property of an image that is shared across different viewers. We introduce a database for which we have measured the probability that each picture will be remembered after a single view. We analyze some of the image features that contribute to making an image memorable, and we train a predictor based on global image descriptors.
We show that predicting image memorability is a task that can be addressed with current computer vision techniques, including global images features used for scene recognition tasks. This work is a first attempt to quantify this useful quality of individual images, with tremendeous applications in multiples fields of fundamental and applied sciences.
|