JPH0245815A - Optical neuro-computer - Google Patents

Optical neuro-computer

Info

Publication number
JPH0245815A
JPH0245815A JP19754788A JP19754788A JPH0245815A JP H0245815 A JPH0245815 A JP H0245815A JP 19754788 A JP19754788 A JP 19754788A JP 19754788 A JP19754788 A JP 19754788A JP H0245815 A JPH0245815 A JP H0245815A
Authority
JP
Japan
Prior art keywords
light
tij
signal
light emitting
element array
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
JP19754788A
Other languages
Japanese (ja)
Other versions
JPH077315B2 (en
Inventor
Kazuo Hisama
和生 久間
Shuichi Tai
田井 修市
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mitsubishi Electric Corp
Original Assignee
Mitsubishi Electric Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mitsubishi Electric Corp filed Critical Mitsubishi Electric Corp
Priority to JP19754788A priority Critical patent/JPH077315B2/en
Priority to GB8915308A priority patent/GB2220780B/en
Priority to DE3922129A priority patent/DE3922129C2/en
Priority to US07/375,813 priority patent/US5095459A/en
Publication of JPH0245815A publication Critical patent/JPH0245815A/en
Publication of JPH077315B2 publication Critical patent/JPH077315B2/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Abstract

PURPOSE:To enable an optical system to be one system by converting the transmitted light of a space optical modulator to an electric signal and after that, separating the signal. CONSTITUTION:Now, a space optical modulator 10 is modulated simultaneously to a positive value Tij<(+)> of a Tij matrix and a negative value Tij<(-)> by angular frequencies omega1 and omega2 respectively. Namely, it can be expressed that [Tij<(+)>cosomega1 t+Tij<(-)>cosomega2t]. When a (j)-th row of the space optical modulator 10 is irradiat ed with a (j)-th light emitting diode 1 and the light of an (i)-th row in the space optical modulator 10 is received in an (i)-th photo-detecting element 3, a signal is obtained so as to be Ui=SIGMATijVj=SIGMA(Tij<(+)>Vjcocomega1t+Tij<(-)>Vjcosomega2t). This signal is passed through a band pass filter 11 and the components of the angular frequencies omega1 and omega2 are separated. After that, the signal is rectified by a rectifier 12 and it is obtained that Ui<(+)>=SIGMATij<(+)>Vj, Ui<(-)>=SIGMATijTij<(-)>Vj. These signals are subtracted by a differential amplifier 4 and threshold processing is executed by a comparator 5. Then, the signals are fed back to the light emitting diode 1.

Description

【発明の詳細な説明】 〔産業上の利用分野〕 この発明は、生物の神経回路網を模倣し、連想機能、パ
ターン認識機能などを有するコンピュータにューロコン
ピュータ)を光波術ヲ用イテ達成スる光ニューロコンピ
ュータに関するものである。
[Detailed Description of the Invention] [Field of Industrial Application] This invention imitates the neural network of living things and achieves the use of light wave technology in computers (eurocomputers) having associative functions, pattern recognition functions, etc. This is related to a light neurocomputer.

〔従来の技術〕[Conventional technology]

第3図は例えば刊行物(電子情報通信学会光・量子エレ
クトロニクス研究会資料0QE87−174(1988
))に示された従来の光連想ニューロコンピュータの構
成図である。図において、  (1a)。
Figure 3 shows, for example, a publication (IEICE Optical and Quantum Electronics Study Group Material 0QE87-174 (1988
)) is a configuration diagram of a conventional light-associative neurocomputer shown in FIG. In the figure (1a).

(1b)Fi発光ダイオードアレイ、  (2a)、 
(2b)は光学マスク、(5a) 、  (5b)は受
光素子プレイ、(4)は差動増幅器、(5)は比較器、
(6)は不完全情報入力。
(1b) Fi light emitting diode array, (2a),
(2b) is an optical mask, (5a) and (5b) are light receiving element plays, (4) is a differential amplifier, (5) is a comparator,
(6) is incomplete information input.

+71は完全情報出力である。+71 is complete information output.

次に動作について説明する。第3図は光連想ニューロコ
ンピュータの基本構成を示したものである。発光ダイオ
ードプレイ(1a)、(lb)  から扇状の光を光学
マスク(2a)、(2b)へ照射する。発光ダイオード
プレイ(1a)、  (1b)はそれぞれの要素が点燈
するか清澄するかに対応して、1か0の状態を表わす一
種のベクトルV=(Vl、 V2.・・・vi。
Next, the operation will be explained. FIG. 3 shows the basic configuration of a light-associative neurocomputer. Light-emitting diode plays (1a) and (lb) irradiate fan-shaped light onto optical masks (2a) and (2b). The light emitting diode play (1a), (1b) is a kind of vector V=(Vl, V2...vi) representing the state of 1 or 0, depending on whether each element is lit or clear.

vn)を表わしている。光学マスク(2a)、  (2
b)はnxrlの要素に分けられておシ、その各要素の
光透過率が異なっている一種のマ) IJクスT=(T
ij)を表わし、受光素子アレイ(3b)、  (sb
)は一種ノヘクト/l/U:=(Ul、 U2. ””
Ul−・・・Un)を表わしている。1番目の発光ダイ
オードで光学マスクのj行目を照射し、光学マスクのi
列目の光をi番目を受光素子で受けると。
vn). Optical mask (2a), (2
b) is divided into nxrl elements, and each element has a different light transmittance.
ij), and the light receiving element array (3b), (sb
) is a kind of nohect/l/U:=(Ul, U2. ””
Ul-...Un). The first light emitting diode illuminates the j-th row of the optical mask, and the i-th row of the optical mask
When the i-th light receiving element receives the light in the column.

Ui =  ΣTijVj なるベクトル、マトリクス演算が行なえる。Ui = ΣTijVj Vector and matrix operations can be performed.

ニューロコンピュータでハ各ニューロン間ノ結合強度に
情報を蓄積している。この光連想ニューロコンピュータ
では光学マスクの透過藁Tに情報を蓄積している。その
情報蓄積規則はHop fieldモデルに基いたもの
であり。
A neurocomputer stores information on the connection strength between each neuron. In this light-associative neurocomputer, information is stored in the transparent straw T of the optical mask. The information storage rules are based on the Hop field model.

Tij :  Σ(2yIsl −t ) (2yj(
s)1 )で与えられる。Tijは正の値も負の値もと
シ得るが、光学的に負の値は扱えないため、第3図に示
すようにTijの成分の正値TiJ (+1と負値Ti
j(−1に対応して光学系を2系統(Ui(+l  U
If−) )作製し、その差 Ui : IJi(+1−Ui(−1 を差動増幅器(4)で得ている。この出力信号を比較器
(5)でしきい値処理 vi=θ(Ui) 但し。
Tij : Σ(2yIsl −t ) (2yj(
s) is given by 1). Tij can take both positive and negative values, but since negative values cannot be handled optically, the positive value TiJ (+1 and negative value Ti
Two optical systems (Ui(+l U
If-)) is produced, and the difference Ui: IJi(+1-Ui(-1) is obtained by the differential amplifier (4). This output signal is subjected to threshold processing by the comparator (5) as vi=θ(Ui ) however.

した後1発光ダイオードアレイ(1a)、  (1b)
にフィードバックしている。
After 1 light emitting diode array (1a), (1b)
are providing feedback.

このようにすると0例えば光学マスク(2a) 。In this way, 0, for example, the optical mask (2a).

(2b)にアルファベットA、J、Hに相当する3種類
の情報を蓄積しておけば、最初に発光ダイオード(1a
)、  (+b)に不完全な情報0例えばA′ を入力
しても、フィードバックを繰り返すうちK。
If three types of information corresponding to alphabets A, J, and H are stored in (2b), the light emitting diode (1a
), (+b) even if you input incomplete information 0, for example A', K after repeating feedback.

その出力は入力情報A′に最も近い情報Aを完全出力と
して表示する。
The output displays the information A closest to the input information A' as the complete output.

このことを別の言葉で言えば、系のエネルギが蓄積情報
A、  J、  Eで極小値を取り、不完全情報が与え
られると、その近くのエネルギ極小値を取るように系全
体が変化して行く(発光ダイオードプレイの点滅状態を
変化させることによる)という事である。
To put this in other words, when the energy of the system takes a minimum value with accumulated information A, J, and E, and incomplete information is given, the entire system changes so that the energy near that value takes a minimum value. (by changing the blinking state of the light emitting diode play).

〔発明が解決しようとするiIl!題〕従来の光ニュー
ロコンピュータは以上のように構成されているので、 
Tijマトリクスの正値、負値に対応して光学系を2系
統必要とし、系が複雑であるという問題があった。
[The problem that the invention attempts to solve! Problem] Since the conventional optical neurocomputer is configured as described above,
There is a problem in that two optical systems are required to correspond to the positive and negative values of the Tij matrix, and the system is complicated.

この発明は上記のような問題点を解消するためになされ
たもので、光学系を1系統で構成できる光ニューロコン
ピュータを得ることを目的とする。
The present invention was made to solve the above-mentioned problems, and an object of the present invention is to obtain an optical neurocomputer that can have a single optical system.

〔課題を解決するための手段〕[Means to solve the problem]

この発明に係る光ニューロコンピュータは、ニューロン
の状態ベクトルに応じて点滅する発光素子アレイ、記憶
情報の正値、負値に対してそれぞれ異なった周波数もし
くは位相、または上記正値。
The optical neurocomputer according to the present invention includes a light emitting element array that blinks according to the state vector of neurons, a frequency or phase that is different for positive values and negative values of stored information, or the above-mentioned positive values.

負値に対して交互の時分割により多重化して駆動するこ
とにより相関行列を実現し、上記発光素子アレイからの
光を変調させて上記ニューロンの状態ベクトルと相関行
列との積信号光を透過する空間光変調器、この空間光変
調器からの透過光を受信し上記積信号の和を得る受光素
子プレイ、この受光素子アレイの出力信号を弁別して上
記正値。
A correlation matrix is realized by multiplexing and driving negative values by alternating time division, and the light from the light emitting element array is modulated to transmit the product signal light of the state vector of the neuron and the correlation matrix. A spatial light modulator, a light-receiving element that receives the transmitted light from the spatial light modulator and obtains the sum of the product signals, and discriminates the output signal of the light-receiving element array to obtain the positive value.

負値に相当する信号を得る信号処理器、この信号処理器
で得られた正値、負値に相当する信号の差を得る差動増
幅器、およびこの差動増幅器で得られた差の値をしきい
値処理して上記発光素子アレイにフィードバックする比
較器を備えたものである。
A signal processor that obtains a signal corresponding to a negative value, a differential amplifier that obtains the difference between the positive value obtained by this signal processor, a signal corresponding to a negative value, and a value of the difference obtained by this differential amplifier. It is equipped with a comparator that performs threshold processing and feeds back to the light emitting element array.

〔作用〕[Effect]

この発明における空間光変調器は、 Tijマトリクス
の正値、負値が多重化されて表現されているので、空間
光変調器の透過光を電気信号に変換した後0分離するこ
とKより、光学系は一系統でよい。
In the spatial light modulator of this invention, the positive and negative values of the Tij matrix are expressed in a multiplexed manner. One system is sufficient.

〔実施例〕〔Example〕

以下、この発明の一実施例を図について説明する。第1
図において、(1)は発光素子アレイすなわち発光ダイ
オードアレイ、(3)は受光素子アレイ。
An embodiment of the present invention will be described below with reference to the drawings. 1st
In the figure, (1) is a light emitting element array, that is, a light emitting diode array, and (3) is a light receiving element array.

(4)は差動増幅器、(5)は比較器、α0は空間光変
調器。
(4) is a differential amplifier, (5) is a comparator, and α0 is a spatial light modulator.

allは信号処理器すなわちバンドパスフィルタ、α2
は整流器、α3は空間光変調器のドライバである。
all is a signal processor, that is, a bandpass filter, α2
is a rectifier, and α3 is a driver for a spatial light modulator.

次に動作について説明する。基本原理は従来例と同じで
ある。本実施例では光学マスクの代わりに、外部信号に
て変調可能な空間光変調器a1を用いている。いま、空
間光変調器αeをT11マトリクスの正値Ti」(+)
と負値Ti j (−1に対して、それぞれ角周波数ω
1.ω2で同時に変調する。すなわち。
Next, the operation will be explained. The basic principle is the same as the conventional example. In this embodiment, a spatial light modulator a1 that can be modulated by an external signal is used instead of an optical mask. Now, the spatial light modulator αe is set to the positive value Ti of the T11 matrix (+)
and the negative value Ti j (-1, respectively, the angular frequency ω
1. Simultaneously modulate at ω2. Namely.

と表わす。j番目の発光ダイオード(1)で空間光変調
器αGのj行目を照射し、空間光変調器a〔のi列目の
光をi番目の受光素子(3)で受けるとIJi = !
’ Ti j Vj 」 なる信号が得られる。この信号をバンドパスフィルタf
i9を通して、角周波数ω1とω2の成分を分離した後
、整流器α2で整流する。
It is expressed as When the j-th light emitting diode (1) illuminates the j-th row of the spatial light modulator αG, and the i-th light receiving element (3) receives the light from the i-th column of the spatial light modulator a, IJi = !
A signal 'Ti j Vj' is obtained. This signal is filtered through a bandpass filter f
After separating the components of angular frequencies ω1 and ω2 through i9, they are rectified by a rectifier α2.

すなわち。Namely.

U、(+l == 、 Tij(+l vJなる信号が
得られる。これらの信号を差動増幅器(41で引算して 、 (4”l   、 (−1 Ui=Ut   Ul なる信号を得、これを比較器(5)でしきい値処理し。
The signals U, (+l == , Tij (+l vJ) are obtained. These signals are subtracted by the differential amplifier (41) to obtain the signal (4"l, (-1 Ui=Ut Ul), which is is subjected to threshold processing using a comparator (5).

発光ダイオードアレイ(1)へフィードバックする。Feedback to the light emitting diode array (1).

なお、上記実施例では、空間光変調器(I[Iを周波数
多重化信号で駆動するものを示したが、同一周波数で位
相の異なる信号で0例えばTijの正値は、 、 f+
1 T13  CQSωt、負値はTij’−’sinωt
のように駆動しても良い。このとき、  ’l”ijの
正値、負値は位相によって弁別できる(位相変調法)。
In the above embodiment, the spatial light modulator (I [I) is driven by a frequency multiplexed signal, but if the signals have the same frequency but different phases, for example, a positive value of Tij is: , f+
1 T13 CQSωt, negative value is Tij'-'sinωt
It may also be driven as follows. At this time, positive values and negative values of 'l''ij can be distinguished by phase (phase modulation method).

また9時分開学重化してもよい。その例を第2図に示す
Also, the school may open at 9 o'clock. An example is shown in FIG.

図において0口)は発光ダイオードプレイ、(3)は受
光素子アレイ、(4)は差動増幅器、(5)は比較器。
In the figure, (0) is a light emitting diode playback, (3) is a light receiving element array, (4) is a differential amplifier, and (5) is a comparator.

noは空間光変調器、(14a)(14b)は空間光変
調器ドライバ、  (1sa) (15b)はパルス発
生器、(16a)(16b)Hサンプルホールドアンプ
である。
No is a spatial light modulator, (14a) and (14b) are spatial light modulator drivers, (1sa) and (15b) are pulse generators, and (16a) and (16b) are H sample and hold amplifiers.

空間光変調器ai#:tそのドライバ(14a) (1
4b)を通じて、パルサ(15a) (15b)によっ
てTij(+)とTi】(−)信号で交互にパルス駆動
される。このとき受光素子+31出力を空間光変調器a
1B駆動信号パルスに同期させてサンプルホールドする
ことKよシ。
Spatial light modulator ai#:t its driver (14a) (1
4b), the pulsers (15a) and (15b) alternately pulse drive with Tij(+) and Ti](-) signals. At this time, the output of the light receiving element +31 is sent to the spatial light modulator a.
It is recommended to sample and hold in synchronization with the 1B drive signal pulse.

7ijマトリクスの正値Tij(+1.負値Tij(−
1を分離できる。後の動作は第1図の実施例と同じであ
る。
7ij matrix positive value Tij (+1. negative value Tij (-
1 can be separated. The subsequent operation is the same as the embodiment shown in FIG.

〔発明の効果〕〔Effect of the invention〕

以上のように、この発明によれば、ニューロンの状態ベ
クトルに応じて点滅する発光素子アレイ。
As described above, according to the present invention, there is provided a light emitting element array that blinks in accordance with the state vector of a neuron.

記憶情報の正値、負値に対してそれぞれ異なった周波数
もしくは位相、または上記正値、負値に対して薬互の時
分割により多重化して駆動することにより相関行列を実
現し、上記発光素子プレイからの光を変調させて上記ニ
ューロンの状態ベクトルと相関行列との積信号光を透過
する空間光変調器、この空間光変調器からの透過光を受
信し上記積信号の和を得る受光素子アレイ、この受光素
子アレイの出力信号を弁別して上記正値、負値に相当す
る信号を得る信号処理器、この信号処理器で得られた正
値、負値に相当する信号の差を得る差動増幅器、および
この差動増幅器で得られた差の値をしきい値処理して上
記発光素子アレイにフィードバックする比較器を備えた
ので、光学系が一系統でよく、装置が簡単となり、大規
模化が容易となる効果がある、
A correlation matrix is realized by driving the positive values and negative values of the stored information at different frequencies or phases, or by multiplexing and driving the positive values and negative values by mutual time division, and the light emitting element A spatial light modulator that modulates the light from the play and transmits the product signal light of the state vector of the neuron and the correlation matrix, and a light receiving element that receives the transmitted light from the spatial light modulator and obtains the sum of the product signals. array, a signal processor that discriminates the output signal of this light receiving element array to obtain signals corresponding to the positive and negative values, and a difference that obtains the difference between the signals corresponding to the positive and negative values obtained by this signal processor. Since it is equipped with a dynamic amplifier and a comparator that thresholds the difference value obtained by this differential amplifier and feeds it back to the light emitting element array, only one optical system is required, which simplifies the device and reduces the size of the device. This has the effect of facilitating scale-up.

【図面の簡単な説明】[Brief explanation of the drawing]

第1図はこの発明の一実施例による光ニューロコンピュ
ータを示す構成図、第2図はこの発明の他の実施例によ
る光ニューロコンピユータラ示ス構成図、第3図は従来
の光ニューロコンピュータを示す構成図である、 図において、 ill、  (1a) 、(1b)は発
光ダイオードアレイ、  (2a)、  (2b)は光
学−qXり、 tag、  (5a)。 (5b)は受光素子アレイ、(4)は差動増幅器、(5
)は比較器、(61は不完全情報入力、(7)は完全情
報出力。 aGは空間光変調器、allはバンドパスフィルタ、a
りは整流器、 +13.  (14a)、  (14b
)はドライバ@  (15a)。 (15b)Fiパルス発生器、  (16a)、 (1
6b)はサンプルホールドアンプである。 なお、各図中同一符号は同一または相当部分を示すもの
とする。
FIG. 1 is a block diagram showing an optical neurocomputer according to an embodiment of the present invention, FIG. 2 is a block diagram showing an optical neurocomputer according to another embodiment of the present invention, and FIG. 3 is a block diagram showing a conventional optical neurocomputer. In the figure, ill, (1a) and (1b) are light emitting diode arrays, (2a) and (2b) are optical-qX arrays, and tag, (5a). (5b) is a photodetector array, (4) is a differential amplifier, (5
) is a comparator, (61 is an incomplete information input, (7) is a complete information output. aG is a spatial light modulator, all is a bandpass filter, a
ri is a rectifier, +13. (14a), (14b
) is driver @ (15a). (15b) Fi pulse generator, (16a), (1
6b) is a sample and hold amplifier. Note that the same reference numerals in each figure indicate the same or corresponding parts.

Claims (1)

【特許請求の範囲】 ニューロンの状態ベクトルと相関行列との積和演算を行
う光ニューロコンピュータにおいて、上記ニューロンの
状態ベクトルに応じて点滅する発光素子アレイ、 記憶情報の正値、負値に対してそれぞれ異なつた周波数
もしくは位相、または上記正値、負値に対して交互の時
分割により多重化して駆動することにより上記相関行列
を実現し、上記発光素子アレイからの光を変調させて上
記ニューロンの状態ベクトルと相関行列との積信号光を
透過する空間光変調器、 この空間光変調器からの透過光を受信し上記積信号の和
を得る受光素子アレイ、 この受光素子アレイの出力信号を弁別して上記正値、負
値に相当する信号を得る信号処理器、この信号処理器で
得られた正値、負値に相当する信号の差を得る差動増幅
器、 およびこの差動増幅器で得られた差の値をしきい値処理
して上記発光素子アレイにフィードバックする比較器を
備えたことを特徴とする光ニューロコンピュータ。
[Claims] An optical neurocomputer that performs a product-sum operation of a neuron state vector and a correlation matrix, comprising: a light emitting element array that blinks in accordance with the neuron state vector; and positive and negative values of stored information. The correlation matrix is realized by multiplexing and driving the positive and negative values at different frequencies or phases, or by alternating time division, and the light from the light emitting element array is modulated to generate the neuron. a spatial light modulator that transmits a product signal light of a state vector and a correlation matrix; a light receiving element array that receives the transmitted light from this spatial light modulator and obtains the sum of the product signals; and a valve that receives the output signal of this light receiving element array. A signal processor that separately obtains signals corresponding to the positive and negative values, a differential amplifier that obtains the difference between the signals corresponding to the positive and negative values obtained by this signal processor, and An optical neurocomputer comprising a comparator that performs threshold processing on the difference value and feeds it back to the light emitting element array.
JP19754788A 1988-07-05 1988-08-08 Optical euro computer Expired - Lifetime JPH077315B2 (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
JP19754788A JPH077315B2 (en) 1988-08-08 1988-08-08 Optical euro computer
GB8915308A GB2220780B (en) 1988-07-05 1989-07-04 Neurocomputer
DE3922129A DE3922129C2 (en) 1988-07-05 1989-07-05 Neurocomputer
US07/375,813 US5095459A (en) 1988-07-05 1989-07-05 Optical neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP19754788A JPH077315B2 (en) 1988-08-08 1988-08-08 Optical euro computer

Publications (2)

Publication Number Publication Date
JPH0245815A true JPH0245815A (en) 1990-02-15
JPH077315B2 JPH077315B2 (en) 1995-01-30

Family

ID=16376298

Family Applications (1)

Application Number Title Priority Date Filing Date
JP19754788A Expired - Lifetime JPH077315B2 (en) 1988-07-05 1988-08-08 Optical euro computer

Country Status (1)

Country Link
JP (1) JPH077315B2 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5386310A (en) * 1991-12-06 1995-01-31 Hitachi, Ltd. Parallel optical switching apparatus
JP2022543366A (en) * 2019-07-29 2022-10-12 ライトマター インコーポレイテッド Systems and methods for analog computing using linear photonic processors

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5386310A (en) * 1991-12-06 1995-01-31 Hitachi, Ltd. Parallel optical switching apparatus
JP2022543366A (en) * 2019-07-29 2022-10-12 ライトマター インコーポレイテッド Systems and methods for analog computing using linear photonic processors
US12244354B2 (en) 2019-07-29 2025-03-04 Lightmatter, Inc. Systems and methods for analog computing using a linear photonic processor

Also Published As

Publication number Publication date
JPH077315B2 (en) 1995-01-30

Similar Documents

Publication Publication Date Title
US5095459A (en) Optical neural network
US20020168100A1 (en) Spatial image processor
CN109450550A (en) A kind of relevant OAM communication demodulating system based on convolutional neural networks
JPH0245815A (en) Optical neuro-computer
RU2057363C1 (en) Hybrid transputer for optoelectronic neural processor
Glover An optical Fourier/electronic neurocomputer automated inspection system
CN117454949A (en) All-optical convolutional neural network device
AP787A (en) Amplitude to phase conversion logic using interference.
JPH03276137A (en) Optical neural processor
CN116882470A (en) Photon convolution neural network chip
JPH06161586A (en) Optical neural processor
US5704015A (en) Optical operation element, optical data processing circuit and photoelectric operation element
JPH06295351A (en) Neural network with lateral interconnection
JPH02125362A (en) Operation system for neurocomputer
JPH06130444A (en) Optical operator element and optical information processing circuit
JP2540200B2 (en) Optical Neuro Computer
CN115759221B (en) A multi-task diffraction photoelectric neural network system and its usage method
US20250371332A1 (en) Three-dimensional programmable neural networks
CN114839789B (en) Diffraction focusing method and device based on binarization spatial modulation
CN116312803B (en) Gene analysis method and device based on optical pulse neural network
Xiao et al. Compatible Learning for Deep Photonic Neural Network
JPH06259157A (en) Learning system for self-organization pattern
JPH0344717A (en) Nerve cell simulating circuit
JPH01286085A (en) Optical learning and recognition machine
JPH022537A (en) Optical learning and recognizing machine