Memorability Prediction
Memorability prediction focuses on understanding and quantifying how well different types of media (images, videos, music, advertisements) are remembered. Current research employs deep learning models, including convolutional neural networks (CNNs) and vision transformers (ViTs), to predict memorability based on features extracted from the media itself, sometimes incorporating additional data like EEG brainwave patterns or textual descriptions. This field is significant for its potential applications in areas such as advertising, user experience design, and educational technology, as well as for advancing our understanding of human memory and perception.
Papers
Overview of the EEG Pilot Subtask at MediaEval 2021: Predicting Media Memorability
Lorin Sweeney, Ana Matran-Fernandez, Sebastian Halder, Alba G. Seco de Herrera, Alan Smeaton, Graham Healy
Predicting Media Memorability: Comparing Visual, Textual and Auditory Features
Lorin Sweeney, Graham Healy, Alan F. Smeaton