Abstract
Purpose: Classification image studies reveal that the neural mechanisms driving perception and saccades during search use information about the target but are also based on an inhibitory surround not present in the target luminance profile (Ludwig et al., 2007; Eckstein et al., 2007). Here, we ask whether these inhibitory surrounds might reflect a strategy that the brain has adapted to optimize the search for targets in natural scenes. To test this hypothesis, we sought to estimate the best linear template (behavioral receptive field), built from linear combinations of Gabor functions representing V1 simple cells in search for an Gaussian target added to natural images. Methods: Statistically non-stationary and non-Gaussian properties of natural scenes preclude calculation of the best linear template from analytic expressions and require an iterative method such as a genetic algorithm (virtual evolution). Thus, we virtually evolved a behavioral receptive field built from linear combinations of Gabor receptive fields to maximize accuracy detecting the Gaussian target in one 4000 calibrated images (van Hateren & van der Schaaf, 1998). Results: We found the optimized linear template included a substantial inhibitory surround that was larger than that found in humans performing target search in white noise (Eckstein et al., 2007). Inclusion of independent internal noise to each channel during the virtual evolution resulted in an optimized template with inhibitory surrounds that were comparable to those in human observers. Finally, the inhibitory surrounds were robust to changes in the contrast of the signal and non-linearities in the model, and generalized to tasks in which the signal occluded other objects in the image. Conclusion: Together the results suggest that the apparent sub-optimality of inhibitory surrounds in human behavioral receptive fields when searching for a target in white noise might reflect a strategy to optimize detection of targets in natural scenes.