A model of motion adaptation and motion after-effects based upon principal component regression

Area of Science:

• Computational Neuroscience
• Visual Perception
• Image Processing

Background:

• Established models of motion detection often use linear regression.
• These models do not fully explain adaptation effects on perceived speed.
• Adaptation to visual stimuli can alter motion perception.

Purpose of the Study:

• To propose a new computational model for motion detection based on principal component regression.
• To explain adaptation effects in motion perception, including perceived speed changes.
• To compare the proposed model with existing linear regression models.

Main Methods:

• Developed a computational model assuming errors in both spatial and temporal dimensions of image signals.
• Utilized principal component regression instead of linear regression.
• Simulated neuron responses and compared model predictions with psychophysical data.

Main Results:

• The proposed model successfully predicts increases and decreases in perceived speed after adaptation to various motion patterns.
• Response suppression in model complex cell neurons accounts for observed adaptation effects.
• Predictions (2) and (3) regarding speed perception changes were not explained by linear regression models.

Conclusions:

• The visual system likely processes motion signals with an inherent assumption of spatial and temporal measurement errors.
• Principal component regression offers a more robust framework for modeling motion perception than linear regression.
• The new model provides a better explanation for the complex effects of adaptation on perceived speed.