Meaning guides attention during scene viewing, even when it is irrelevant

Area of Science:

• Cognitive psychology
• Neuroscience
• Computer vision

Background:

• Human visual attention must prioritize regions in complex scenes.
• The interplay between low-level image salience and high-level semantic meaning in attentional guidance is debated.
• Prior research suggested meaning dominates attention when directly contrasted.

Purpose of the Study:

• To investigate the role of semantic meaning in attentional guidance when visual salience is task-relevant.
• To compare the influence of meaning and salience on attention using controlled visual stimuli.
• To test whether meaning guides attention even when irrelevant to the task.

Main Methods:

• Utilized meaning maps and saliency maps to represent semantic and salience features, respectively.
• Designed brightness rating and search tasks where salience was relevant, but meaning was not.
• Measured attentional distribution using fixation density maps to compare meaning and salience effects.

Main Results:

• Meaningful scene regions received higher attentional priority than visually salient regions.
• This prioritization occurred even in tasks where semantic information was irrelevant.
• Fixation patterns indicated that meaning consistently guided attention over salience.

Conclusions:

• Scene semantics play a dominant role in guiding human visual attention.
• Meaningful content appears to be prioritized automatically, irrespective of task demands.
• Findings support theories emphasizing the fundamental role of semantic processing in visual attention.