| CPC G06T 7/74 (2017.01) [G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] | 20 Claims |
|
1. A method of configuring an artificial neural network to be selective to curvature, wherein the artificial neuron network comprises a first topographically organized layer of orientation-selective neurons, and wherein each orientation-selective neuron of a subset of the orientation-selective neurons is selective to an oriented line of inputs in an image patch, and wherein the subset of the orientation-selective neurons is collectively selective of a plurality of orientations in the image patch, the method comprising:
creating a second topographically organized layer of neurons selective for curve segments, wherein each curve-segment-selective neuron of a subset of the curve-segment-selective neurons is selective to a curve segment, and wherein the subset of curve-segment-selective neurons is collectively selective of a plurality of curve segments in the image patch; in which each curve-segment-selective neuron is configured to respond to a set of line segments in the image patch by selection of inputs from orientation-selective neurons from the first topographically organized layer, wherein excitatory weights are configured for selected inputs that are selective for line segments that form the curve segment and have positions and orientations that match the curve segment, with inputs of other orientations and locations configured to have less weight including inhibition; and
creating a curve-selective neuron, wherein the curve-selective neuron is selective to a curve having a specified center, a specified degree of curvature, and a specified orientation, the curve-selective neuron having as input an output of the topographically organized layer of curve-segment-selective neurons, in which the curve-selective neuron responds to the specified degree of curvature and at the specified orientation relative to the specified center by selection of inputs from curve-segment-selective neurons having an orientation that is determined systematically based on the position of the input in relation to the center, and wherein the selection is further based on a correspondence between the specified degree of curvature and a corresponding property for which individual input curve-segment-selective neurons are selective.
|