Contour features predict valence and threat judgements in scenes

Area of Science:

• Visual perception
• Environmental psychology
• Survival behavior

Background:

• Rapidly assessing environmental threats is crucial for survival.
• Scene gist extraction from visual scenes may aid threat detection.

Purpose of the Study:

• To investigate the relationship between visual scene features and emotional judgments.
• To determine how specific image statistics correlate with perceived safety or threat.

Main Methods:

• Extracted curvature, length, and orientation statistics from the International Affective Picture System (IAPS) image set.
• Correlated these statistics with emotional valence scores.
• Created abstract line drawings manipulating these features and collected participant ratings for positivity/negativity and safety/threat.

Main Results:

• Images with angular contours were associated with negative ratings.
• Images with long contours were associated with positive ratings.
• Abstract scenes with smooth, long, horizontal contours were rated as positive/safe.
• Abstract scenes with short, angular contours were rated as negative/threatening.

Conclusions:

• Specific combinations of visual features (length, angularity, orientation) significantly influence judgments of environmental threat.
• These findings highlight a mechanism by which visual scene analysis contributes to survival behaviors.