Speed and direction of locally-paired dot patterns

Area of Science:

• Visual Perception
• Computational Neuroscience
• Motion Perception

Background:

• Phenomenal transparency in random-dot kinematograms is abolished under specific 'locally-balanced' conditions.
• Previous research attributed this abolition to local inhibition between motion detectors with large directional differences.

Purpose of the Study:

• To investigate perceived motion in locally-balanced random-dot kinematograms with varying directional differences.
• To determine if local interactions abolish global motion perception or merely alter its characteristics.

Main Methods:

• Subjects performed direction and speed judgments on locally-paired stimuli with component directions 60, 90, or 120 degrees apart.
• Stimuli included variations in dot lifetime and component velocities.

Main Results:

• Subjects consistently perceived coherent motion and made reliable direction judgments, even with large directional differences.
• Judged motion direction and speed adhered to a vector-average rule for both same- and mixed-velocity components.
• The abolition of transparency did not equate to the abolition of a global motion percept.

Conclusions:

• Local interactions that abolish transparency generate a vector combination of superimposed motions, not solely inhibitory.
• Perceived motion in these displays is a vector average of component motions, indicating a global integration process.
• Findings challenge the notion that local inhibition is the sole mechanism underlying the loss of transparency.