Spatiotemporal interpolation and quality of apparent motion

Area of Science:

• Visual perception
• Neuroscience
• Computational vision

Background:

• Apparent motion is perceived when discrete stimuli are presented sequentially.
• Understanding the visual system's ability to process motion is crucial for visual neuroscience.
• Distinguishing between smooth and sampled motion perception informs models of visual processing.

Purpose of the Study:

• To determine the conditions under which apparent motion is indistinguishable from smooth motion.
• To compare the discrimination of smooth versus apparent motion with the visual system's interpolation precision.
• To investigate the influence of stimulus parameters on motion perception.

Main Methods:

• Observers discriminated between smooth and apparent motion under varying step sizes, contrasts, velocities, and stimulus types (lines, bars, gratings).
• Thresholds for smooth versus sampled motion discrimination were measured.
• Precision of apparent motion interpolation was assessed by measuring thresholds for discriminating Vernier offsets along motion trajectories.
• The effect of motion speed and inter-station distance on interpolation accuracy was evaluated.

Main Results:

• Discrimination thresholds between smooth and apparent motion were around 40 arc minutes under optimal conditions.
• Tolerated step size increased with velocity, particularly for low-spatial-frequency stimuli.
• Tolerated step size decreased with longer presentation durations and higher stimulus contrast.
• Precise visual interpolation was achieved for inter-station distances as small as 2 arc minutes.

Conclusions:

• The visual system exhibits distinct capabilities for discriminating smooth from apparent motion and for interpolating motion.
• A spatiotemporal filtering model can effectively explain the observed results in both discrimination and interpolation tasks.
• These findings provide insights into the early stages of visual motion processing.