We seek to understand how simple patterns are preceived by humans, and how human psychophysical performance at discriminating between two patterns can be related, in a quantitative manner, to activity in small networks of visual neurons.
To this end, we measured contrast detection, contrast increment, contrast masking, orientation discrimination and spatial frequency discrimination thresholds for spatially localized stimuli at 4deg of eccentricity. Our stimulus geometry emphasizes interactions among overlapping visual filters and differs from that used in previous threshold measurements, which also admits interactions between distant filters. We quantitatively account for all measurements by simulating a small population of overlapping visual filters interacting through divisive inhibition. We depart from previous models of this kind in the parameters of divisive inhibition and in using a statistically efficient decision stage based on Fisher information. The success of this unified account suggests that, contrary to Bowne (1990), spatial vision thresholds reflect a single level of processing, perhaps as early as primary visual cortex.
We used five pattern discrimination tasks to investigate early visual processing in humans: contrast discrimination, orienation discrimination, spatial period discrimination, contrast masking with variable mask orientation, and contrast masking with variable mask spatial period.
A simple computational model of just one hypercolumn in primary visual cortex was used to quantitatively relate psychophysical thresholds measured in humand to activity at the level of single neurons.
Remarkably, it was possible to predict all human data with a single set of model parameters, although the nature of the tasks studied varied widely. Furthermore, the model parameters derived from predicting the human psychophysical data matched well the known characteristics of biological neurons in primary visual cortex.
For more information, see our publications page, and in particular our recent article in The Journal of the Optical Society of America [A].
Copyright © 2000 by the University of Southern California, iLab and Prof. Laurent Itti