View-adaptation reveals coding of face pose along image, not object, axes

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Psychology

Background:

• High-level adaptation effects offer insights into neural coding for objects and faces.
• View adaptation is crucial for understanding face representation and view-invariant recognition.
• The study explores whether depth rotations of faces are encoded based on the image or the object itself.

Purpose of the Study:

• To determine if depth rotations of faces are represented in an image-based or object-based manner.
• To dissociate object-based axes from image-based axes using a 90° planar rotation of an adapting face.
• To analyze participants' responses pre- and post-adaptation to infer the representational format.

Main Methods:

• Utilized a view adaptation paradigm.
• Employed a 90° planar rotation of the adapting face to separate image-based and object-based axes.
• Collected and analyzed participants' responses before and after adaptation.

Main Results:

• Participant responses were most consistent with an image-based representation of facial depth rotations.
• The findings indicate that the visual system encodes depth rotations relative to the image plane.
• This suggests a reliance on image-specific features rather than abstract object-based properties for this aspect of face processing.

Conclusions:

• The neural representation of facial depth rotations appears to be predominantly image-based.
• Planar rotation effectively dissociated image-based from object-based representations in this context.
• These results contribute to understanding the mechanisms underlying face perception and view-invariant recognition.