Mid-level perceptual features contain early cues to animacy.

Area of Science:

• Cognitive Neuroscience
• Visual Perception
• Object Recognition

Background:

• Limited research exists on how the visual system makes large-scale object distinctions like animacy and size.
• Neural representations reflect object size and animacy, with distinct mid-level perceptual features for different sizes.

Purpose of the Study:

• To test if animate and manmade objects differ in mid-level perceptual features.
• To investigate the role of these features in early visual processing and object classification.

Main Methods:

• Generated synthetic images ('texforms') preserving texture and form but not basic-level identification.
• Utilized visual search efficiency to measure perceptual similarity.
• Employed electrophysiological methods to track processing stages.

Main Results:

• Faster visual search for animals among manmade objects than among other animals (and vice versa), even with unrecognizable texforms.
• This mixed-animacy search advantage originates in early target individuation stages.
• Perceived curvature significantly contributes to the animacy-based search advantage and classification.

Conclusions:

• Mid-level perceptual features, particularly curvature, provide cues for animacy classification.
• The visual system leverages these early perceptual cues for efficient object detection, recognition, and classification.