Lossy processing principles in 2D and 3D vision

Area of Science:

• Cognitive Science
• Neuroscience
• Computer Vision

Background:

• The human visual system processes vast amounts of information, yet perception is often simplified.
• Similarities exist between 2D image statistics and 3D visual perception.
• Understanding these similarities can reveal fundamental principles of visual processing.

Purpose of the Study:

• To explore the common principles underlying 2D summary statistics and 3D visual processing.
• To investigate how information is discarded in visual perception.
• To explain visual illusions such as change blindness as a consequence of information loss.

Main Methods:

• Comparative analysis of 2D and 3D visual processing mechanisms.
• Information theory applied to visual perception models.
• Psychophysical experiments investigating visual illusions and awareness.

Main Results:

• 2D summary statistics and fragmentary 3D vision share common information processing principles.
• Both 2D and 3D visual systems discard redundant or inessential information for ecologically valid perception.
• Information loss, when consistency assumptions are violated, leads to phenomena like change blindness without conscious awareness.

Conclusions:

• Visual processing prioritizes essential information, discarding redundancy for efficiency.
• The principles of information discarding are fundamental to both 2D and 3D vision.
• Visual illusions highlight the active, assumption-driven nature of perception.