Skip to main content
Center for Open Science

Rethinking Vividness: Semantic moreso than perceptual features drive the recall vividness of complex pictures

Preprints
Morales-Torres, R; Davis, S; Cabeza, R
September 25, 2024

While some memories are vivid and detailed, others are vague and indistinct. Although this is a ubiquitous experience, the cognitive mechanisms underlying these differences are not well understood. To investigate this issue, in Experiment 1 (N=122), we examined how the subjective feeling of memory vividness relates to the perceptual (i.e., color and brightness) and semantic (i.e., the category of a stimulus) properties of naturalistic images. Surprisingly, our findings showed that while perceptual properties moderately contribute to the vividness of visual memories, the primary driver is semantic properties. Moreover, in Experiment 2 (N=132), by employing different Deep Neural Networks (DNNs) to select stimuli based on their semantic features, we could reliably identify stimuli that generate vivid memories. Our results demonstrate, at multiple levels of inference, that although vivid memories are often compared to perception, this subjective quality of memory is strongly influenced by our knowledge of the world.

Duke Scholars

Altmetric Attention Stats
Dimensions Citation Stats

DOI

Publication Date

September 25, 2024
 

Citation

APA
Chicago
ICMJE
MLA
NLM
Morales-Torres, R., Davis, S., & Cabeza, R. (2024). Rethinking Vividness: Semantic moreso than perceptual features drive the recall vividness of complex pictures. Center for Open Science. https://doi.org/10.31219/osf.io/cvb4t
Morales-Torres, Ricardo, Simon Davis, and Roberto Cabeza. “Rethinking Vividness: Semantic moreso than perceptual features drive the recall vividness of complex pictures.” Center for Open Science, September 25, 2024. https://doi.org/10.31219/osf.io/cvb4t.
Morales-Torres, Ricardo, et al. “Rethinking Vividness: Semantic moreso than perceptual features drive the recall vividness of complex pictures.” Center for Open Science, 25 Sept. 2024. Crossref, doi:10.31219/osf.io/cvb4t.

DOI

Publication Date

September 25, 2024