🤖 AI Summary
This study addresses intra-class variability in image memorability—the differential likelihood of remembering individual instances within the same semantic category—a phenomenon previously uncharacterized in computational vision.
Method: We introduce “intra-class memorability” as a novel construct and develop a quantitative framework grounded in human behavioral experiments (serial delayed matching), cognitive psychological measurements, and statistical modeling. We define and implement ICMscore, a fine-grained, time-interval–aware metric that captures memorability dynamics across repeated exposures.
Contribution/Results: We identify local texture, edge contrast, and spatial distribution of semantically salient regions as key visual determinants of intra-class memorability. This work constitutes the first systematic computational model of memorability at the intra-class level, providing empirically validated insights and a publicly available computational tool. It advances robustness in vision models, informs principled training-data curation, and supports the development of human–machine collaborative cognitive architectures.
📝 Abstract
We introduce intra-class memorability, where certain images within the same class are more memorable than others despite shared category characteristics. To investigate what features make one object instance more memorable than others, we design and conduct human behavior experiments, where participants are shown a series of images one at a time, and they must identify when the current item matches the item presented a few steps back in the sequence. To quantify memorability, we propose the Intra-Class Memorability score (ICMscore), a novel metric that incorporates the temporal intervals between repeated image presentations into its calculation. Our contributions open new pathways in understanding intra-class memorability by scrutinizing fine-grained visual features that result in the least and most memorable images and laying the groundwork for real-world applications in cognitive science and computer vision.