FamFac - A Database of Famous Faces for Psychology Experiments

Area of Science:

• Cognitive Neuroscience
• Psychology
• Computer Vision

Background:

• Previous studies on face recognition were potentially biased by variations in low-level visual properties (e.g., brightness, resolution) and familiarity levels of famous faces.
• These uncontrolled variables can affect event-related potential (ERP) detection and complicate the disentanglement of familiar and unfamiliar face processing.

Purpose of the Study:

• To develop a freely available database of famous faces with homogenized low-level visual properties and established familiarity levels.
• To provide a standardized resource that mitigates confounding effects in face perception research.

Main Methods:

• A custom algorithm was developed to standardize image brightness.
• Image size and resolution were homogenized using GIMP software.
• Familiarity levels for 183 famous faces were assessed by 48 Portuguese college students.

Main Results:

• The custom algorithm successfully standardized image brightness, with individual images not significantly differing from the mean.
• Forty-one faces were identified as highly familiar within the dataset.
• The database provides a controlled set of stimuli for face processing research.

Conclusions:

• The developed algorithm and database are freely available, offering valuable resources for the scientific community.
• This standardized database helps to avoid confounding factors present in previous face recognition studies.
• The resource facilitates more reliable investigations into the neural mechanisms of familiar and unfamiliar face recognition.