Sensors Know Which Photos Are Memorable

The goal of this study is to determine if physiological signals are salient in the detection of memorable personal photos. We begin by collecting physiological sensor data as well as memorability and emotion ratings for photos. We then build a mixed model to evaluate the predictive power of physiological variables on memorability and emotion by examining whether or not the photographer's data is useful for predicting the ratings of the photographer or the ratings of the subjects in the photos. Our results suggest that heart rate and GSR (galvanic skin response) data are the major predictors of memorability for photographers, and that the sensor signals are not particularly useful for predicting memorability ratings of subjects in the photos.

[1]  J. D. McGaugh,et al.  A Novel Demonstration of Enhanced Memory Associated with Emotional Arousal , 1995, Consciousness and Cognition.

[2]  Bernard Ghanem,et al.  What Makes an Object Memorable? , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[3]  Chuan Qin,et al.  TagSense: a smartphone-based approach to automatic image tagging , 2011, MobiSys '11.

[4]  Da Young Ju,et al.  1G-36 Memorable Moment Extraction using the Pupil Dilation Analysis , 2013 .

[5]  Kai Kunze,et al.  Personalized record of the city wander with a wearable device: a pilot study , 2016, UbiComp Adjunct.

[6]  Jianxiong Xiao,et al.  Memorability of Image Regions , 2012, NIPS.

[7]  Antonio Torralba,et al.  Modifying the Memorability of Face Photographs , 2013, 2013 IEEE International Conference on Computer Vision.

[8]  Antonio Torralba,et al.  Understanding the Intrinsic Memorability of Images , 2011, NIPS.

[9]  Kristina Höök,et al.  AffectCam: arousal- augmented sensecam for richer recall of episodic memories , 2013, CHI Extended Abstracts.

[10]  James L. McGaugh,et al.  Make mild moments memorable: add a little arousal , 2006, Trends in Cognitive Sciences.

[11]  Antonio Torralba,et al.  Understanding and Predicting Image Memorability at a Large Scale , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).