Direction discrimination for band-pass filtered random dot kinematograms

Area of Science:

• Visual perception
• Computational neuroscience
• Psychophysics

Background:

• Apparent motion perception is limited by a maximum spatial displacement (dmax).
• Previous research indicated a monotonic increase in errors with displacement.
• The role of stimulus characteristics, like spatial frequency, on dmax is not fully understood.

Purpose of the Study:

• To investigate apparent motion direction discrimination using spatially band-pass filtered random dots.
• To analyze how stimulus center frequency (Fc) affects the spatial limit (dmax).
• To re-evaluate the relationship between displacement, error, and dmax.

Main Methods:

• Direction discrimination tasks were performed using spatially band-pass filtered random dot stimuli.
• Stimulus displacements were varied relative to the stimulus center frequency (Fc).
• Performance was analyzed in relation to the stimulus autocorrelation function.

Main Results:

• Direction discrimination errors did not show a monotonic increase with displacement for filtered stimuli.
• Performance oscillated around chance level for displacements exceeding 1 cycle of Fc.
• Systematic errors in perceived direction were observed at approximately 1 1/4 cycles of Fc.
• The upper spatial limit (dmax) was defined on the initial rising portion of the displacement-error function.
• dmax was found to scale inversely with Fc, consistent with previous findings but extending to higher spatial frequencies.

Conclusions:

• The relationship between displacement and motion discrimination error is complex and frequency-dependent.
• The definition of dmax as the initial rising portion of the displacement-error function is robust.
• The inverse scaling of dmax with Fc holds for a wider range of spatial frequencies than previously reported.