EXPRESS: Anchor Objects Contribute to Scene Function Understanding

Area of Science:

• Cognitive Psychology
• Visual Perception
• Scene Understanding

Background:

• Everyday actions focus on objects, yet scene context also influences understanding.
• The role of object-level versus scene-level information in scene function comprehension is debated.

Purpose of the Study:

• To investigate whether actions shape scene hierarchy and if object-level information is key to scene function understanding.
• To determine if action-related anchor objects impact scene function comprehension.

Main Methods:

• Experiment 1: Participants matched action words to scenes with masked action-related anchors (REL), unrelated anchors (UNREL), or random objects (RAND).
• Experiment 2: Lexical decision task using scenes as primes for action words to measure scene function understanding.
• Control experiment: Assessed if anchor absence impaired scene categorization.

Main Results:

• Scene matching performance was significantly worse in REL scenes compared to UNREL and RAND scenes.
• The REL condition showed significantly slower response times in the lexical decision task than the RAND condition.
• Results indicate that scene function understanding primarily depends on action-related object-level information.

Conclusions:

• Scene function understanding is mainly driven by action-related anchor objects.
• Scene context alone is not always sufficient for understanding a scene's functions.
• Visual realism and object clutter influence performance but not the core findings.