US5355437A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Neural network architecture for pattern recognition
Assignee: Yozan Inc
•
Filing: 1993-02-16
•
Granted: 1994-10-11
•
Expires: 2013-02-16
Abstract
Abstract A data processing system according to the present invention provides a plurality of neural layers with neuron groups, each neuron group having a fixed number of neurons. The neurons of the neuron group each have an output coupled to a neuron of an adjacent neuron layer. Each neuron layer has a plurality of neuron groups, and each neuron group has at least one neuron which also belongs to another neuron group, resulting in an overlap in the neuron groups. The number of neurons in the nth neural layer is determined on the basis of the number of neurons in the (n-1)th layer, the size of the neuron groups, and the degree of overlap between the adjacent neuron groups. In a variation of the data processing system of the present invention, the data processing system comprises a plurality of mutually independent data processing portions, each of which comprises a plurality of neural layers.
Inventors
- 👤 Sunao Takatori
- 👤 Makoto Yamamoto
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-28 05:45
CPC Classifications
G; G06; G06N; G06N3/00; G06N3/02; G06N3/04; G06V; G06V30/00; G06V30/10; G06V30/18; G06V30/1801; G06V30/18019; G06V30/18038; G06V30/18048; G06V30/18057; G06N3/045; G06N3/0464; G06N3/0495; G06N3/08; G06N3/09; G06V10/00; G06V10/70; G06V10/82; G06V30/19; G06V30/191
Claims Analysis
Total Claims: 3
Claim 1
Independent Claim
248 words
A data processing system for recognizing an input image in response to an input characteristics data signal and outputting a recognition result signal, the data processing system comprising: at least first, second and third adjacent square neural layers, each of said neural layers comprising a plurality of neurons arranged in neuron groups, each neuron group of said each neural layer having a predetermined number "a" of neurons, said predetermined number of neurons being coupled to a neuron in an adjacent neural layer, said neuron of said adjacent neural layer being responsive to said neurons of said neuron group; wherein each neuron group of said each neural layer has at least one neuron which also belongs to an adjacent neuron group of said each neural layer so that said neuron group and said adjacent neuron group overlap by a width "v" of neurons, said overlap width "v" having a value of at least one and being less than a width of said neuron groups, said value "a" being equal for each of said first, second and third neural layers; and wherein one of said at least first, second and third neural layers outputs said recognition result signal, each of said neural layers having an arrangement such that an nth neural layer has N n number of neurons such that: ##EQU6## the term "N n-1 " being a number of neurons in an adjacent (n-1)th layer, said arrangement causing said data processing system to have a minimum number of neurons.
Claim 2
Dependent Claim
248 words
A data processing system for recognizing an input image in response to an input characteristics data signal and outputting a recognition result signal, the data processing system comprising: at least first, second and third adjacent square neural layers, each of said neural layers comprising a plurality of neurons arranged in neuron groups, each neuron group of said each neural layer having a predetermined number "a" of neurons, said predetermined number of neurons being coupled to a neuron in an adjacent neural layer, said neuron of said adjacent neural layer being responsive to said neurons of said neuron group; wherein each neuron group of said each neural layer has at least one neuron which also belongs to an adjacent neuron group of said each neural layer so that said neuron group and said adjacent neuron group overlap by a width "v" of neurons, said overlap width "v" having a value of at least one and being less than a width of said neuron groups, said value "a" being equal for each of said first, second and third neural layers; and wherein one of said at least first, second and third neural layers outputs said recognition result signal, each of said neural layers having an arrangement such that an nth neural layer has N n number of neurons such that: ##EQU6## the term "N n-1 " being a number of neurons in an adjacent (n-1)th layer, said arrangement causing said data processing system to have a minimum number of neurons.
Claim 3
Dependent Claim
128 words
A data processing system as recited in claim 1, wherein: said at least first, second and third neural layers cooperate to form preprocessing means for generating clustering results in response to said input characteristics data signal, said preprocessing means including mutually exclusive independent data processing portions for outputting a corresponding clustering result; and said data processing system further comprising post-processing means for outputting said recognition result signal in response to said clustering results from said preprocessing means, said post-processing means including an input neural layer and a second plurality of neural layers having said arrangement between said nth and said (n-1)th neural layers each of said independent data processing portions of said preprocessing means outputting said corresponding clustering results to said input neural layer of said post-processing means.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1997-10-10 | 1998-04-11 | 1998-10-11 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2001-10-10 | 2002-04-11 | 2002-10-11 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2005-10-10 | 2006-04-11 | 2006-10-11 | $7700.00 | Window Closed / Lapsed |