Scene context automatically drives predictions of object transformations.

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Computer Vision

Background:

• Human object perception is context-dependent, aiding recognition.
• Previous research focused on context disambiguating static object properties.
• The dynamic updating of object representations based on scene context remains under-investigated.

Purpose of the Study:

• To investigate if human observers dynamically predict object appearance changes with viewpoint shifts.
• To determine if scene orientation automatically influences object representations.
• To test the automaticity of scene-based predictions in object perception.

Main Methods:

• Three behavioral experiments involving 152 participants.
• Temporarily occluding objects within rotating scenes.
• Assessing performance on an orthogonal perceptual discrimination task upon object reappearance.

Main Results:

• Perceptual task performance was significantly affected by the coherence of object reappearance with scene rotation.
• This scene-based prediction effect persisted despite a majority of trials violating real-world contingencies.
• Findings demonstrate automaticity in scene-based predictive processing.

Conclusions:

• Contextual information plays a crucial role in predicting object transformations in real-world environments.
• Human visual system automatically updates object representations based on surrounding scene orientation.
• Scene-based predictions are a fundamental aspect of object perception and transformation processing.