The Representational Similarity between Visual Perception and Recent Perceptual History.

Area of Science:

• Cognitive Neuroscience
• Visual Perception
• Neuroimaging

Background:

• Perceptual stability despite fluctuating visual input is crucial for reliable object recognition.
• Serial dependence, a bias towards recently viewed stimuli, may explain perceptual stability by smoothing noisy sensory information.
• The neural mechanisms underlying serial dependence remain incompletely understood.

Purpose of the Study:

• To investigate the neural underpinnings of serial dependence in visual object perception.
• To determine if neural representations of past stimuli influence current object processing.
• To examine whether this influence involves shared neural codes between memory and perception.

Main Methods:

• Electroencephalography (EEG) was used to record brain responses in human observers.
• Observers viewed prototypical objects (faces, cars, houses) and morphs.
• Representational Similarity Analysis (RSA) was applied to EEG data.

Main Results:

• Behavioral responses showed a bias towards previously seen objects, confirming serial dependence.
• EEG data revealed that neural responses to current objects contained information about the preceding stimulus.
• This influence of past stimuli on current neural representations did not involve identical neural codes.

Conclusions:

• Neural activity reflects the influence of recent visual history on current perception.
• Serial dependence arises from a pre-processing brain state rather than direct representational overlap.
• Past stimulus history influences current representations without implying a shared neural code between memory and perception.