Signal-tuned Gabor functions as models for stimulus-dependent cortical receptive fields

Area of Science:

• Computational Neuroscience
• Neuroscience
• Visual Cortex Modeling

Background:

• V1 cells are fundamental to visual processing.
• Existing models struggle to fully capture V1 cell receptive field properties and stimulus-dependent responses.
• Gabor functions offer a promising framework for modeling neural receptive fields.

Purpose of the Study:

• To propose and analyze a novel model for V1 cell receptive fields and responses.
• To utilize signal-tuned Gabor functions and the signal-tuned Gabor transform (STGT) for V1 cell modeling.
• To replicate key properties and stimulus-dependent behaviors of V1 cells.

Main Methods:

• Modeling V1 cell receptive fields using signal-tuned Gabor functions.
• Expressing V1 cell responses using the signal-tuned Gabor transform (STGT).
• Testing the model with standard grating and slit visual stimuli.

Main Results:

• The proposed model accurately replicates V1 cell properties with standard inputs.
• The model successfully emulates the stimulus-dependent character of V1 cells.
• Signal-tuned Gabor functions provide exact representations of tuning signals.

Conclusions:

• The signal-tuned Gabor function model offers a robust framework for understanding V1 cell function.
• This model advances the accurate detection of spatial and spectral events in visual processing.
• The findings align with recent neurophysiological studies on V1 cell responses.