Right place, right time: Spatiotemporal predictions guide attention in dynamic visual search

Area of Science:

• Cognitive Psychology
• Neuroscience
• Human-Computer Interaction

Background:

• Visual search is typically studied in static lab settings, but real-world search occurs in dynamic environments.
• Incorporating the temporal dimension is crucial for understanding real-life visual search behavior.

Purpose of the Study:

• To investigate how humans learn and utilize spatiotemporal regularities in dynamic environments.
• To assess the impact of predictable versus unpredictable target appearances on visual search performance.

Main Methods:

• A dynamic-search task was designed with targets fading in and out amidst distractors.
• Participants searched for targets with predictable spatiotemporal patterns versus unpredictable ones.
• Eye-movement recordings were used to analyze attention guidance.

Main Results:

• Performance was significantly better for spatiotemporally predictable targets compared to unpredictable ones.
• Learning effects were primarily long-term, though single-trial priming also contributed.
• Eye-tracking data indicated that spatiotemporal regularities proactively guide attention.

Conclusions:

• Spatiotemporal regularities across space and time effectively guide visual search.
• This guidance is largely driven by robust, long-term learned representations of environmental regularities.