Artificial neurons using delta-sigma modulation
First Claim
1. A neuron comprising:
- a plurality of input lines for providing a plurality of input signals to said neuron;
a neuron input unit receiving said input signals and outputting an intermediate signal;
a delta sigma modulator Which includes a saturating adder, said delta sigma modulator receiving said intermediate signal from said neuron input unit and producing a neuron output signal in response.
0 Assignments
0 Petitions
Accused Products
Abstract
An artificial neuron for use in a neural processing network comprises a plurality of input signal lines, an arrangement for computing a nonlinear function of the sum of the inputs multiplied by associated weights, and a saturating delta-sigma modulator which oversamples the computed value and produces an encoded neuron output signal. Conversion of signals for use by these neurons preferably is performed by delta-sigma modulators at the inputs to the neurons, which may be incorporated directly into sensors. Processing of the output signals from the neuron includes low-pass filtering and decimation. The present invention may be used in many diverse areas. For example, arrays of sensors with delta signal modulators may be coupled with a network of the neurons to form an intelligent vision system. Linear signal processing, both conventional and adaptive, can be done by a simple neuronal system that operates linearly.
41 Citations
44 Claims
-
1. A neuron comprising:
-
a plurality of input lines for providing a plurality of input signals to said neuron; a neuron input unit receiving said input signals and outputting an intermediate signal; a delta sigma modulator Which includes a saturating adder, said delta sigma modulator receiving said intermediate signal from said neuron input unit and producing a neuron output signal in response. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22)
-
-
23. A neuron comprising:
-
a plurality of input lines for providing a plurality of input signals to said neuron; a neuron input unit receiving said input signals and outputting an intermediate signal; a delta sigma modulator receiving said intermediate signal from said neuron input unit and producing a neuron output signal in response; wherein said neuron input unit includes a selector connected to each of said plurality of input lines to select a weight-product corresponding to each of said plurality of input signals. - View Dependent Claims (24)
-
-
25. A neuron for a neural processing network comprising:
-
a plurality of input lines for providing a plurality of input signals to said neuron; a neuron input unit receiving said input signals and outputting an intermediate signal; a delta sigma modulator receiving said intermediate signal from said neuron input unit and producing a neuron output signal in response; wherein said delta sigma modulator includes a saturating adder receiving the intermediate signal from said neuron input unit and also receiving feedback through the neuron output signal. - View Dependent Claims (26, 27, 28)
-
-
29. An artificial neuron receiving a plurality of input signals on a plurality of input signal lines, comprising:
-
a selector unit connected to each of said plurality of input lines, said selector unit selecting and applying a weighting element to each of said plurality of input signals to produce a weight-product on each of said input lines; an adder for computing the sum of all weight-products and producing an intermediate output signal; and a delta sigma modulator receiving said intermediate output signal and producing a neuron output signal. - View Dependent Claims (30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40)
-
-
41. An adaptive linear combiner, comprising:
-
a delta sigma modulator for inputting a signal and providing an output signal; a shift register receiving the output signal from said delta sigma modulator and outputting a plurality of output signals; a delta sigma neuron connected to the shift register for receiving the plurality of output signals and providing an output signal; and a low pass filter and decimator connected to the delta sigma neuron and receiving the output signal from the delta sigma neuron. - View Dependent Claims (42, 43, 44)
-
Specification