Human visual search follows a suboptimal Bayesian strategy revealed by a spatiotemporal computational model and...

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Vision Science

Background:

• Conflicting evidence exists on the optimality of human eye movements during visual search.
• Previous models often neglect fixation duration and memory limitations, leading to discrepancies with empirical data.

Purpose of the Study:

• To investigate the spatial and temporal properties of human eye movements during visual search.
• To develop a continuous-time eye movement model incorporating saccadic inaccuracy, bias, and memory constraints.

Main Methods:

• Measured the temporal dynamics of human visibility maps.
• Conducted a visual search experiment.
• Developed and validated a computational model of eye movement control.

Main Results:

• The developed model shows improved agreement with spatial and temporal aspects of human eye movements.
• Predicted human memory capacity is approximately eight previous fixations.
• Demonstrated that human eye movement strategies are suboptimal, likely for cost-efficiency.

Conclusions:

• Human visual search eye movements are not spatially optimal but rather a cost-effective strategy.
• The model provides a better account of human eye movement metrics, including fixation duration and memory.
• Findings contribute to understanding the complex interplay of perception, memory, and motor control in visual search.