Testing neuronal accounts of anisotropic motion perception with computational modelling

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Neurons in early visual cortex show over-representation for cardinal orientations and directions.
• Neuronal anisotropies are hypothesized to explain perceptual phenomena like the oblique effect and reference repulsion.
• It's unclear if these anisotropies can simultaneously explain both perceptual effects.

Purpose of the Study:

• To investigate if neuronal anisotropies can simultaneously account for the oblique effect and reference repulsion in motion perception.
• To determine the dependence of these effects on stimulus duration.
• To explore the relationship between adaptation, perceptual effects, and neuronal anisotropies.

Main Methods:

• Conducted psychophysical experiments measuring motion perception.
• Manipulated stimulus duration and used brief adaptation protocols.
• Implemented and tested various neuronal decoding models with varied anisotropy parameters.

Main Results:

• The oblique effect and reference repulsion were independent of stimulus duration.
• Brief adaptation to a single direction induced reference repulsion in the orientation domain and the inverse oblique effect in the direction domain.
• No tested neuronal decoding model could fully explain the observed perceptual data.

Conclusions:

• The oblique effect likely stems from anisotropic preferred direction distributions in V1 and MT.
• Reference repulsion appears to be a separate phenomenon, possibly originating in higher-order cortical areas.
• A unified explanation for both the oblique effect and reference repulsion based solely on early visual cortical anisotropies is unlikely.