US5361328A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Data processing system using a neural network
Assignee: Ezel Inc
•
Filing: 1993-05-07
•
Granted: 1994-11-01
•
Expires: 2013-05-07
Abstract
Abstract A data processing system comprising a plurality of neural layers characterized in that a part of neurons in one of the layers are connected to neurons in the following layer.
Inventors
- 👤 Sunao Takatori
- 👤 Ryohei Kumagai
- 👤 Koji Matsumoto
- 👤 Makoto Yamamoto
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-06 10:39
CPC Classifications
G; G06; G06N; G06N3/00; G06N3/02; G06N3/04; G06N3/0464; G06N3/0495
Claims Analysis
Total Claims: 8
Claim 1
Independent Claim
202 words
A data processing system comprising: a plurality of neural layers each including a plurality of neurons from a first layer to at least one intermediate layer and a last neural layer, said first layer receiving an input data and processing said input data to transmit a processed result to the at least one intermediate layer, each intermediate layer receiving data processed by a previous layer to process said data and to output a data as a processed result, said last neural layer connected to receive data processed by previous layer so as to process said data and to output an output data, wherein a plurality of groups of neurons are defined for each neural layer and a plurality of hexagonal-shaped neural cells including and plurality of neural layers are defined by said groups, and wherein said input data is image data which is divided into a plurality of local areas of a hexagonal shape; and means for performing hexagonal convolution image processing between said hexagonal shape input data and said plurality of hexagonal-shaped neural cells, further comprising a neural network connected to said last neural layer for receiving said output data so as to extract a resultant data from said output data.
Claim 2
Dependent Claim
202 words
A data processing system comprising: a plurality of neural layers each including a plurality of neurons from a first layer to at least one intermediate layer and a last neural layer, said first layer receiving an input data and processing said input data to transmit a processed result to the at least one intermediate layer, each intermediate layer receiving data processed by a previous layer to process said data and to output a data as a processed result, said last neural layer connected to receive data processed by previous layer so as to process said data and to output an output data, wherein a plurality of groups of neurons are defined for each neural layer and a plurality of hexagonal-shaped neural cells including and plurality of neural layers are defined by said groups, and wherein said input data is image data which is divided into a plurality of local areas of a hexagonal shape; and means for performing hexagonal convolution image processing between said hexagonal shape input data and said plurality of hexagonal-shaped neural cells, further comprising a neural network connected to said last neural layer for receiving said output data so as to extract a resultant data from said output data.
Claim 3
Dependent Claim
30 words
A data processing system as claimed in claim 1, further comprising a neural network connected to said first neural layer for inputting said input data to said first neural layer.
Claim 4
Dependent Claim
34 words
A data processing system as claimed in claim 1, wherein data volume of said output data of each said neural cells is equal to data volume of said output data of other neural layers.
Claim 5
Dependent Claim
36 words
A data processing system as claimed in claim 3, wherein said neural network comprises a first neural layer having a plurality of neurons each of which receives all of said output data of said last layer.
Claim 6
Dependent Claim
92 words
A data processing system as claimed in claim 4, wherein each said plurality of neurons receiving said output data correspond to one of said neural cells, and are connected with all of the neural cells with a connection weight, wherein a connection weight value is a maximum value for the connection weight connected to the corresponding neural cell, has a negative value for the connection weight connected to the neural cells adjacent to said corresponding neural cell and increases in value for neural cells as distance from said corresponding neural cell increases.
Claim 7
Dependent Claim
197 words
A data processing system comprising: a plurality of neural layers each including a plurality of neurons from a first layer to at least one intermediate layer and a last neural layer, said first layer receiving an input data and processing said input data to transmit a processed result to the at least one intermediate layer, each intermediate layer receiving data processed by a previous layer to process said data and to output a data as a processed result, said last neural layer connected to receive data processed by a previous layer so as to process said data and to output an output data, wherein a plurality of groups of neurons are defined for each neural layer and a plurality of hexagonal-shaped neural cells including and plurality of neural layers are defined by said groups, and wherein said input data is image data which is divided into a plurality of local areas of a hexagonal shape; and means for performing hexagonal convolution image processing between said hexagonal shape input data and said plurality of hexagonal-shaped neural cells, further comprising a neural network connected to said first neural layer for inputting said input data to said first neural layer.
Claim 8
Dependent Claim
197 words
A data processing system comprising: a plurality of neural layers each including a plurality of neurons from a first layer to at least one intermediate layer and a last neural layer, said first layer receiving an input data and processing said input data to transmit a processed result to the at least one intermediate layer, each intermediate layer receiving data processed by a previous layer to process said data and to output a data as a processed result, said last neural layer connected to receive data processed by a previous layer so as to process said data and to output an output data, wherein a plurality of groups of neurons are defined for each neural layer and a plurality of hexagonal-shaped neural cells including and plurality of neural layers are defined by said groups, and wherein said input data is image data which is divided into a plurality of local areas of a hexagonal shape; and means for performing hexagonal convolution image processing between said hexagonal shape input data and said plurality of hexagonal-shaped neural cells, further comprising a neural network connected to said first neural layer for inputting said input data to said first neural layer.
References Cited (Backward Citations)
US3097349A
Cited
US3250918A
Cited
US3165644A
Cited
US3287649A
Cited
US3284780A
Cited
US3311895A
Cited
US3317901A
Cited
US3317900A
Cited
US3333249A
Cited
US3533072A
Cited
US3691400A
Cited
US3602888A
Cited
US3601811A
Cited
US3599009A
Cited
US3568075A
Cited
US3638196A
Cited
US3701974A
Cited
US3950733A
Cited
US4044243A
Cited
US4163983A
Cited
US4254474A
Cited
US4319331A
Cited
US4326259A
Cited
US4450530A
Cited
US4774677A
Cited
US4518866A
Cited
US4799149A
Cited
US4716540A
Cited
US4760604A
Cited
US4792915A
Cited
US4739496A
Cited
US4750153A
Cited
US4766568A
Cited
US4803736A
Cited
US4760437A
Cited
US4800519A
Cited
US4737929A
Cited
US4773024A
Cited
US4802103A
Cited
US4792460A
Cited
US4748674A
Cited
US4805225A
Cited
US4752906A
Cited
US4796199A
Cited
US4809193A
Cited
US4807168A
Cited
US4762397A
Cited
US5067164A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1997-10-31 | 1998-05-02 | 1998-11-01 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2001-10-31 | 2002-05-02 | 2002-11-01 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2005-10-31 | 2006-05-02 | 2006-11-01 | $7700.00 | Window Closed / Lapsed |