Object identification is isolated from scene semantic constraint: evidence from object type and token discrimination

Area of Science:

• Cognitive Psychology
• Visual Perception
• Computational Neuroscience

Background:

• Scene perception involves understanding the context provided by a scene.
• Two models, description enhancement and criterion modulation, propose scene context aids object identification.
• These models suggest schema activation facilitates object processing.

Purpose of the Study:

• To test two models of scene-context effects on object identification.
• To investigate whether scene meaning influences object perception.
• To determine if semantic consistency aids or hinders object recognition.

Main Methods:

• Utilized forced-choice type-discrimination and token-discrimination paradigms.
• Presented briefly displayed scenes with semantically consistent and inconsistent objects.
• Measured participants' ability to discriminate between objects based on their semantic relationship to the scene.

Main Results:

• Object discrimination was superior for semantically inconsistent objects compared to consistent ones.
• No performance benefit for semantically consistent objects was found in token-discrimination tasks.
• Results contradicted predictions from both description enhancement and criterion modulation models.

Conclusions:

• Scene meaning does not appear to facilitate initial object perceptual analysis.
• Stored knowledge of real-world scene-object contingencies does not influence object identification.
• Object recognition processes seem independent of semantic scene context.