METHOD FOR EXTRACTING SPATIAL KNOWLEDGE FROM AN INPUT SIGNAL USING COMPUTATIONAL MANIFOLDS

• US 20100316283A1
• Filed: 08/23/2010
• Published: 12/16/2010
• Est. Priority Date: 05/16/2006
• Status: Active Grant

• Current Assignee: Douglas S. Greer
• Original Assignee: Douglas S. Greer
• Inventors: Greer, Douglas S.
• Granted Patent: US 8,041,651 B2
• Time in Patent Office: Days
• US Class Current: 382/155
• CPC Class Codes: G06F 18/21355 nonlinear criteria, e.g. embedding a manifold in a Euclidean space
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology
• CPC Class Codes: G06V 10/7715 Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods

Abstract

A processor architecture for a learning machine is presented which uses a massive array of processing elements having local, recurrent connections to form global associations between functions defined on manifolds. Associations between these functions provide the basis for learning cause-and-effect relationships involving vision, audition, tactile sensation and kinetic motion. Two arbitrary input signals hold each other in place in a manifold association processor and form the basis of short-term memory.

Claims

1. A method for generating a spatially-encoded output signal from an input signal having a structure comprising a plurality of discrete signal units, the method comprising:

inputting the input signal into a computer processor having software stored therein for executing a plurality of processing elements, wherein each processing element comprises a logic manifold association processor having an input and an output, and wherein each processing element generates a single output signal;

transforming the input signal into representations defined on manifolds of one of a physical space, a projective geometry, and topological manifolds of rotations or translations;

forming a recursive loop between a pair of processing elements, wherein the output of each processing element of the pair of processing elements is fed back into the input of the other processing element of the pair, and wherein at least one of the pair of processing elements further receives an element input comprising a small subset of the plurality of discrete signal units of the input signal;

training a plurality of pairs of processing elements to detect and form associations between a plurality of reciprocal signal pairs, wherein when a first signal of a reciprocal signal pair of the plurality is input into a first processing element, the first processing element outputs a second signal of the reciprocal signal pair, and when the second signal of the reciprocal signal pair is input into a second processing element, the second processing element outputs the first signal, so that the reciprocal signal pair converges to a fixed signal that is locked within the pair of processing elements;

wherein when the at least one of the pair of processing elements that receives the small subset of discrete signal units receives the input signal units, the two processing elements generate an encoded output comprising spatial information associated with the subset of input signal units; and

storing the encoded output in a memory in communication with the computer processor.

2. The method of claim 1, wherein the discrete signal units within the input signal have corresponding spatial locations that are encoded with a frequency representation, wherein each processing element encodes a different grid of spatial locations.

3. The method of claim 1, wherein the input signal further comprises temporal knowledge where a product manifold is formed with the spatial manifolds and a time manifold, or a frequency representation of time, the resulting knowledge encoding events corresponding to a unique time and place.

4. The method of claim 1, wherein the input signal is a three-dimensional image and the step of transforming comprises aligning the image with a plane of symmetry.