Local feature suppression effect in face and non-face stimuli

Area of Science:

• Cognitive psychology
• Neuroscience
• Visual perception

Background:

• Human faces capture attention automatically and are processed more efficiently than other stimuli.
• Automatic face processing suggests specialized neural mechanisms, potentially involving feature detectors.
• It remains debated whether face detectors process global configurations or both global and local features.

Purpose of the Study:

• To investigate whether face-tuned visual detectors process global face configurations exclusively or also local face features.
• To determine if local face features exhibit local feature suppression (LFS) compared to non-face stimuli.
• To elucidate the encoding mechanism of faces as basic visual features.

Main Methods:

• Two experiments were conducted using visual search tasks.
• Participants identified the presence or absence of a local target within face (normal/scrambled) and non-face (Kanizsa diamond/house) configurations.
• Reaction times and error rates were measured to assess processing efficiency.

Main Results:

• A robust local feature suppression (LFS) effect was observed for face stimuli, impacting both reaction times and error rates.
• Control stimuli (non-faces) showed either no difference or a local feature enhancement effect.
• These findings indicate differential processing of local features in faces versus non-faces.

Conclusions:

• Faces are encoded by globally tuned face detectors, supporting the concept of faces as a basic visual feature.
• The observed LFS effect suggests that globally tuned detectors are less efficient at processing local face features.
• This research clarifies the nature of face perception and its underlying neural processing mechanisms.