Modelling adaptation to directional motion using the Adelson-Bergen energy sensor.

Area of Science:

• Computational neuroscience
• Visual perception

Background:

• The motion energy sensor is a standard model for human visual system movement sensing.
• Adaptation effects are crucial in motion perception studies but not modeled in current energy sensors.

Purpose of the Study:

• To extend the motion energy sensor model to incorporate adaptation-induced changes.
• To provide a computational model that accounts for adaptation in motion perception.

Main Methods:

• Developed an extended motion energy sensor model.
• Incorporated a leaky integrator (RC gain-control circuit) applied to the sensor's time-dependent output.
• Computed space-time representation of sensor output.

Main Results:

• The extended model demonstrates a decline in sensor output during prolonged stimulation.
• The model shows changes in the relative outputs of different sensors after adaptation.
• Simulated adaptation effects mirror psychophysical observations.

Conclusions:

• The extended motion energy sensor model successfully incorporates adaptation.
• This model provides a framework for understanding adaptation in motion perception computationally.
• The model aligns with psychophysical data on motion adaptation.