OVERVIEW

In this paper, we provide the first attempt to answer the question: what exactly about an image is remembered? To do this, we augmented both the images and object segmentations from the PASCAL-S dataset with ground truth memorability scores and shed light on the various factors and properties that make an object memorable (or forgettable) to humans. In addition, we explore the effectiveness of deep learning and other computational approaches in predicting object memorability in images



DATASET

• Memorability scores for all 3,412 objects and 850 images from the PASCAL-S dataset

• Memorability maps corresponding to each of the 850 PASCAL-S images

• All original 850 PASCAL-S images and object segmentations used for the experiment

Download all files here (.zip).

AUTHORS

Rachit Dubey

Joshua Peterson

Aditya Khosla

Ming-Hsuan Yang

Bernard Ghanem