Facial expression categorization predominantly relies on mid-spatial frequencies.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Vision Science

Background:

• Facial expressions are vital for human communication and social interaction.
• Understanding the role of spatial frequencies (SFs) in emotion perception is an active area of research.
• Previous studies have yielded mixed results regarding the optimal SFs for recognizing basic facial expressions.

Purpose of the Study:

• To investigate the optimal spatial frequencies (SFs) for recognizing basic facial expressions.
• To determine the influence of viewing distance on facial expression recognition.
• To address the gap in understanding SFs' diagnosticity in facial expression processing.

Main Methods:

• Employed two complementary methods: a data-driven approach and a naturalistic simulation of viewing distance variations.
• Experiment 1 used a data-driven method without arbitrary SF cut-offs.
• Experiment 2 simulated naturalistic viewing conditions with varying distances.

Main Results:

• A predominant role of low SFs was observed, but mid-range SFs (approximately 6-13 cycles per face) were found to be crucial for categorization.
• Optimal facial expression recognition occurred at short to medium viewing distances (1.2-2.4 m).
• Recognition performance declined significantly at greater distances, correlating with the loss of mid-range SF information.

Conclusions:

• Facial expression recognition primarily depends on mid-range SF content, which has been previously overlooked.
• Viewing distance critically impacts the availability of necessary SFs for accurate expression recognition.
• Methods excluding mid-SF content provide an incomplete understanding of SFs' role in facial expression processing.