Prediction of uncertain visual trajectories is biased toward motion continuity

Area of Science:

• Perception and Cognitive Science
• Computational Neuroscience
• Visual Perception

Background:

• Visual motion prediction under uncertainty requires integrating statistical and kinematic information.
• Random trajectories present challenges due to noise, necessitating adaptive decision-making strategies.

Purpose of the Study:

• To investigate how decision-making processes and psychophysical parameters are modulated during the extrapolation of random trajectories with different noise characteristics (Random Walk, RDW, and Independently and Identically Distributed, IID).
• To determine the influence of trajectory properties on human visual motion prediction.

Main Methods:

• Participants judged the endpoint of downward-moving dots with varying horizontal noise (RDW or IID).
• Experiments manipulated the last visible position and mean position characteristics.
• Behavioral data and Diffusion Decision Model (DDM) analyses were employed.

Main Results:

• For RDW trajectories, participants accurately used the last visible position, enhancing decision speed and accuracy.
• For IID trajectories, participants showed varied strategies but predominantly focused on the last segment, a suboptimal yet effective approach.
• The perceptual system appears to prioritize smooth motion interpretation for predicting uncertain trajectories.

Conclusions:

• The human visual system favors smooth motion interpretation for predicting uncertain trajectories.
• Focusing on the last visible position is a robust strategy for visual motion prediction, even with non-optimal trajectory types.
• Understanding these prediction mechanisms is crucial for visual perception research.