CN116542305A - Robust optical neural network design method and device for resisting incident signal errors - Google Patents
Robust optical neural network design method and device for resisting incident signal errors Download PDFInfo
- Publication number
- CN116542305A CN116542305A CN202310325450.7A CN202310325450A CN116542305A CN 116542305 A CN116542305 A CN 116542305A CN 202310325450 A CN202310325450 A CN 202310325450A CN 116542305 A CN116542305 A CN 116542305A
- Authority
- CN
- China
- Prior art keywords
- neural network
- optical neural
- optical
- incident signal
- random noise
- 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.)
- Pending
Links
- 230000003287 optical effect Effects 0.000 title claims abstract description 126
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 124
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000013461 design Methods 0.000 title claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 55
- 238000006243 chemical reaction Methods 0.000 claims abstract description 18
- 238000009826 distribution Methods 0.000 claims abstract description 13
- 238000012545 processing Methods 0.000 claims description 22
- 230000006870 function Effects 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 6
- 230000002452 interceptive effect Effects 0.000 claims description 5
- 230000000704 physical effect Effects 0.000 claims description 5
- 230000007704 transition Effects 0.000 claims 1
- 238000004422 calculation algorithm Methods 0.000 abstract description 9
- 230000002093 peripheral effect Effects 0.000 abstract description 8
- 238000005516 engineering process Methods 0.000 abstract description 7
- 238000004364 calculation method Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 238000003062 neural network model Methods 0.000 description 4
- 238000003491 array Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000012938 design process Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G02—OPTICS
- G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
- G02B27/00—Optical systems or apparatus not provided for by any of the groups G02B1/00 - G02B26/00, G02B30/00
- G02B27/0012—Optical design, e.g. procedures, algorithms, optimisation routines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/067—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using optical means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Neurology (AREA)
- Optics & Photonics (AREA)
- Optical Communication System (AREA)
Abstract
The application relates to the technical field of optical neural networks, in particular to a robust optical neural network design method and device for resisting incident signal errors, wherein the method comprises the following steps: acquiring an incident signal error caused by the loading and conversion process of an electrical input signal; determining random noise with a target size range and probability distribution according to the incident signal error, and superposing the random noise on the original optical input signal to obtain an optical input signal based on the random noise; generating a training set by utilizing an optical input signal based on random noise, training the optical neural network by utilizing the training set until training is finished, and obtaining the robust optical neural network with anti-incident signal error. Therefore, the problems that the related technology adopts peripheral light paths and circuits and combines an error calibration algorithm to compensate errors such as phase and amplitude of light, and the like, and the calibration time is long and the difficulty is high are solved.
Description
Technical Field
The present disclosure relates to the field of optical neural networks, and in particular, to a method and an apparatus for designing a robust optical neural network with resistance to error of an incident signal.
Background
The light has the advantages of the fastest propagation speed of the physical space and multidimensional and multi-scale, the light replaces electrons with light, the circuit is replaced by light paths, and the optical computing chip has subversion advantages of high speed, parallelism, low power consumption and the like. Particularly, with the deep development of artificial intelligence algorithms, the mathematical expression of the physical process of limited light propagation in a medium has high similarity with the deep neural network algorithm, and the realization of the neural network calculation by adopting an optical chip is expected to break through the energy efficiency bottleneck of the traditional electronic chip.
The input signals of the optical neural network chip generally need to load the electric signals to the physical characteristics such as amplitude phase polarization and the like of the optical signals through devices such as modulators, phase shifters and the like, errors are inevitably caused in the loading and converting processes of the signals, and the accuracy of the optical neural network chip consistent with that of model training in the reasoning process cannot be ensured.
In the related technology, peripheral light paths and circuits are usually adopted after chip manufacturing is completed and errors such as phases and amplitudes of light are compensated by combining an error calibration algorithm, but the calibration time is long and difficult, and the technical route of calibrating each chip one by one cannot meet the requirement of future mass production.
Disclosure of Invention
The application provides a robust optical neural network design method, device, electronic equipment and storage medium for resisting incident signal errors, which are used for solving the problems that in the related technology, errors such as phase and amplitude of light are compensated by adopting a peripheral light path and a circuit and combining an error calibration algorithm, the calibration time is long, the difficulty is high and the like.
An embodiment of a first aspect of the present application provides a method for designing a robust optical neural network against an incident signal error, including the steps of: acquiring an incident signal error caused by the loading and conversion process of an electrical input signal; determining random noise with a target size range and probability distribution according to the incident signal error, and superposing the random noise on an original optical input signal to obtain an optical input signal based on the random noise; generating a training set by using the optical input signal based on random noise, and training the optical neural network by using the training set until training is finished, so as to obtain the robust optical neural network with the function of resisting incident signal errors.
Optionally, in an embodiment of the present application, the random noise is a combined model of one or more physical properties of light, and is used to eliminate errors caused by signal conversion and loading in the calculation and reasoning process of the optical neural network chip.
Optionally, in one embodiment of the present application, after obtaining the robust optical neural network with the anti-incident signal error, the method further includes: determining a network weight parameter according to the robust optical neural network with the anti-incident signal error; and determining processing parameters of the optical neural network chip based on the network weight parameters, and processing the optical neural network chip by utilizing the processing parameters.
Optionally, in an embodiment of the present application, the optical neural network includes any one of a diffractive neural network, an interfering neural network, and a scattering neural network.
Embodiments of a second aspect of the present application provide a robust optical neural network design apparatus against incident signal errors, comprising: the acquisition module is used for acquiring an incident signal error caused by the loading and conversion process of the electrical input signal; the first determining module is used for determining random noise with a target size range and probability distribution according to the incident signal error, and superposing the random noise on an original optical input signal to obtain an optical input signal based on the random noise; and the training module is used for generating a training set based on the optical input signal based on the random noise, and training the optical neural network by using the training set until the training is finished to obtain the robust optical neural network with the function of resisting the incident signal error.
Optionally, in an embodiment of the present application, the random noise is a combined model of one or more physical properties of light, and is used to eliminate errors caused by signal conversion and loading in the calculation and reasoning process of the optical neural network chip.
Optionally, in one embodiment of the present application, further includes: the second determining module is used for determining a network weight parameter according to the robust optical neural network with the anti-incident signal error after the robust optical neural network with the anti-incident signal error is obtained; and determining processing parameters of the optical neural network chip based on the network weight parameters, and processing the optical neural network chip by utilizing the processing parameters.
Optionally, in an embodiment of the present application, the optical neural network includes any one of a diffractive neural network, an interfering neural network, and a scattering neural network.
An embodiment of a third aspect of the present application provides an electronic device, including: the system comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the robust optical neural network design method for resisting the incident signal error.
An embodiment of a fourth aspect of the present application provides a computer-readable storage medium having stored thereon a computer program that is executed by a processor for implementing a robust optical neural network design method against an incident signal error as described in the above embodiment.
Therefore, the application has at least the following beneficial effects:
according to the method and the device, the target size range and probability distribution of random noise can be determined through the incident signal errors caused by the electric input signal loading and converting process, proper random noise is added to the input light field signals in the training set in the training process of acquiring the neural network weight parameters, errors caused by the signal loading and converting process can be simulated, the signal loading errors are added to the model training in the network model design stage, the performance influence caused by the signal loading and converting is reduced, and the robustness of the optical neural network chip is improved. Therefore, the problems that the related technology adopts peripheral light paths and circuits and combines an error calibration algorithm to compensate errors such as phase and amplitude of light, and the like, and the calibration time is long and the difficulty is high are solved.
Additional aspects and advantages of the application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a flow chart of a method for designing a robust optical neural network against incident signal errors according to an embodiment of the present application;
FIG. 2 is a schematic diagram showing a comparison of training effects between different training methods according to an embodiment of the present application;
fig. 3 is a schematic diagram of a calculation reasoning flow of an optical neural network chip according to an embodiment of the present application;
FIG. 4 is a block diagram of a robust optical neural network design device against incident signal errors, according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and intended for the purpose of explaining the present application and are not to be construed as limiting the present application.
The following describes a method, an apparatus, an electronic device, and a storage medium for designing a robust optical neural network against an incident signal error according to an embodiment of the present application with reference to the accompanying drawings. In order to solve the problems mentioned in the background art, the application provides a robust optical neural network design method for resisting incident signal errors, in the method, the target size range and probability distribution of random noise are determined through the incident signal errors caused by the electric input signal loading and conversion process, proper random noise is added to the light field signals input in the training set in the training process for acquiring the weight parameters of the neural network, errors caused in the signal loading and conversion process can be simulated, and the signal loading errors are added into the model training in the network model design stage, so that the performance influence caused by the signal loading and conversion is reduced, and the robustness of the optical neural network chip is improved. Therefore, the problems that the related technology adopts peripheral light paths and circuits and combines an error calibration algorithm to compensate errors such as phase and amplitude of light, and the like, and the calibration time is long and the difficulty is high are solved.
Specifically, fig. 1 is a schematic flow chart of a robust optical neural network design method for resisting an incident signal error according to an embodiment of the present application.
As shown in fig. 1, the method for designing a robust optical neural network resistant to an incident signal error includes the following steps:
in step S101, an incident signal error caused by the loading and conversion process of the electrical input signal is obtained.
It can be understood that errors exist in the loading and conversion processes of input signals of the optical neural network chip, and the accuracy of the optical neural network chip consistent with that of model training cannot be guaranteed in the reasoning process.
In step S102, random noise having a target size range and probability distribution is determined from the incident signal error, and the random noise is superimposed on the original optical input signal to obtain a random noise-based optical input signal.
In one embodiment of the present application, random noise is a combined model of one or more physical characteristics of light, which is used to eliminate errors caused by signal conversion and loading in the calculation and reasoning process of the optical neural network chip, and may be phase, amplitude, etc., without specific limitation.
It can be understood that the random noise in the embodiment of the application can simulate errors caused in the signal loading and conversion process, so that the errors caused by signal conversion and loading in the optical neural network chip calculation reasoning process are eliminated.
In step S103, a training set is generated using the optical input signal based on random noise, and the optical neural network is trained using the training set until the training is completed, thereby obtaining a robust optical neural network with an anti-incident signal error.
In the design process of the optical neural network, parameters of the optical network structure are required to be obtained through training, as shown in fig. 2, a training set is trained according to the input light field distribution under ideal conditions in the conventional method, and a neural network model meeting the precision requirement under the training set is obtained. In the embodiment of the application, the random noise is added to the training set input light field of the optical neural network model, and the size range and probability distribution of the random noise are determined by errors brought by the signal loading and converting process. Based on the improved training set, a corresponding neural network model meeting the precision requirement is obtained.
In summary, a general workflow for performing reasoning calculation on completing chip processing based on a network model in the embodiment of the application is shown in fig. 3, and the embodiment of the application trains an input signal in advance in a network model training stage to obtain an optimized optical neural network model, so that the influence of random errors introduced by a signal conversion and loading part on the accuracy of the optical neural network chip in the reasoning calculation process can be eliminated, and the robustness of the optical neural network is improved through network model design and training.
It should be noted that the training method for enhancing robustness of the optical neural network in the embodiments of the present application is applicable to different neural network implementations, including but not limited to, a diffraction neural network, an interference neural network, and a scattering neural network.
Optionally, in one embodiment of the present application, after obtaining the robust optical neural network with the anti-incident signal error, the method further includes: determining a network weight parameter according to a robust optical neural network with an anti-incident signal error; and determining processing parameters of the optical neural network chip based on the network weight parameters, and processing the optical neural network chip by using the processing parameters.
The robust optical neural network with the anti-incident signal error can ensure that the optical neural network chip obtains the accuracy consistent with that of model training in the reasoning process, and the processing parameters of the optical neural network chip are determined through the robust optical neural network with the anti-incident signal error so as to realize the processing of the optical neural network chip, so that the robustness of the optical neural network chip in the actual operation process is greatly improved.
According to the robust optical neural network design method for resisting the incident signal errors, the target size range and the probability distribution of random noise are determined through the incident signal errors caused by the electric input signal loading and converting process, proper random noise is added to the input light field signals in the training set in the training process of acquiring the weight parameters of the neural network, errors caused in the signal loading and converting process can be simulated, the signal loading errors are added into the model training in the network model design stage, the performance influence caused by the signal loading and converting is reduced, and the robustness of the optical neural network chip is improved. Therefore, the problems that the related technology adopts peripheral light paths and circuits and combines an error calibration algorithm to compensate errors such as phase and amplitude of light, and the like, and the calibration time is long and the difficulty is high are solved.
Next, a robust optical neural network design apparatus for resisting an incident signal error according to an embodiment of the present application will be described with reference to the accompanying drawings.
Fig. 4 is a block schematic diagram of a robust optical neural network design device for resisting incident signal errors according to an embodiment of the present application.
As shown in fig. 4, the robust optical neural network design apparatus 10 for resisting an incident signal error includes: an acquisition module 100, a first determination module 200, and a training module 300.
The acquisition module 100 is configured to acquire an incident signal error caused by an electrical input signal loading and converting process; the first determining module 200 is configured to determine random noise with a target size range and probability distribution according to an incident signal error, and superimpose the random noise on an original optical input signal to obtain an optical input signal based on the random noise; the training module 300 is configured to generate a training set using the optical input signal based on random noise, and train the optical neural network using the training set until the training is finished, thereby obtaining a robust optical neural network with an anti-incident signal error.
In one embodiment of the present application, random noise is a combined model of one or more physical properties of light, used to eliminate errors due to signal conversion and loading during the computational reasoning of an optical neural network chip.
In one embodiment of the present application, the apparatus 10 of the embodiment of the present application further includes: and a second determination module.
The second determining module is used for determining a network weight parameter according to the robust optical neural network with the anti-incident signal error after the robust optical neural network with the anti-incident signal error is obtained; and determining processing parameters of the optical neural network chip based on the network weight parameters, and processing the optical neural network chip by using the processing parameters.
In one embodiment of the present application, the optical neural network includes any one of a diffractive neural network, an interfering neural network, and a scattering neural network.
It should be noted that the explanation of the embodiment of the method for designing a robust optical neural network against an incident signal error is also applicable to the device for designing a robust optical neural network against an incident signal error of the embodiment, and is not repeated here.
According to the robust optical neural network design device for resisting the incident signal errors, the target size range and the probability distribution of random noise are determined through the incident signal errors caused by the electric input signal loading and converting process, proper random noise is added to the input light field signals in the training set in the training process of acquiring the weight parameters of the neural network, errors caused in the signal loading and converting process can be simulated, the signal loading errors are added into the model training in the network model design stage, the performance influence caused by the signal loading and converting is reduced, and the robustness of the optical neural network chip is improved. Therefore, the problems that the related technology adopts peripheral light paths and circuits and combines an error calibration algorithm to compensate errors such as phase and amplitude of light, and the like, and the calibration time is long and the difficulty is high are solved.
Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device may include:
memory 501, processor 502, and a computer program stored on memory 501 and executable on processor 502.
The processor 502 implements the robust optical neural network design method against incident signal errors provided in the above-described embodiments when executing a program.
Further, the electronic device further includes:
a communication interface 503 for communication between the memory 501 and the processor 502.
Memory 501 for storing a computer program executable on processor 502.
The memory 501 may include high speed RAM (Random Access Memory ) memory, and may also include non-volatile memory, such as at least one disk memory.
If the memory 501, the processor 502, and the communication interface 503 are implemented independently, the communication interface 503, the memory 501, and the processor 502 may be connected to each other via a bus and perform communication with each other. The bus may be an ISA (Industry Standard Architecture ) bus, a PCI (Peripheral Component, external device interconnect) bus, or EISA (Extended Industry Standard Architecture ) bus, among others. The buses may be divided into address buses, data buses, control buses, etc. For ease of illustration, only one thick line is shown in fig. 5, but not only one bus or one type of bus.
Alternatively, in a specific implementation, if the memory 501, the processor 502, and the communication interface 503 are integrated on a chip, the memory 501, the processor 502, and the communication interface 503 may perform communication with each other through internal interfaces.
The processor 502 may be a CPU (Central Processing Unit ) or ASIC (Application Specific Integrated Circuit, application specific integrated circuit) or one or more integrated circuits configured to implement embodiments of the present application.
The embodiments of the present application also provide a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the robust optical neural network design method against incident signal errors as described above.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "N" is at least two, such as two, three, etc., unless explicitly defined otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more N executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.
It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. As with the other embodiments, if implemented in hardware, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable gate arrays, field programmable gate arrays, and the like.
Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, includes one or a combination of the steps of the method embodiments.
Although embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the application, and that variations, modifications, alternatives, and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the application.
Claims (10)
1. The design method of the robust optical neural network for resisting the incident signal error is characterized by comprising the following steps of:
acquiring an incident signal error caused by the loading and conversion process of an electrical input signal;
determining random noise with a target size range and probability distribution according to the incident signal error, and superposing the random noise on an original optical input signal to obtain an optical input signal based on the random noise;
generating a training set by using the optical input signal based on random noise, and training the optical neural network by using the training set until training is finished, so as to obtain the robust optical neural network with the function of resisting incident signal errors.
2. The method of claim 1, wherein the random noise is a combined model of one or more physical properties of light for eliminating errors due to signal conversion and loading during computational reasoning of the optical neural network chip.
3. The method of claim 1, further comprising, after obtaining the robust optical neural network with resistance to incident signal errors:
determining a network weight parameter according to the robust optical neural network with the anti-incident signal error;
and determining processing parameters of the optical neural network chip based on the network weight parameters, and processing the optical neural network chip by utilizing the processing parameters.
4. A method according to any one of claims 1-3, wherein the optical neural network comprises any one of a diffractive neural network, an interfering neural network, and a scattering neural network.
5. A robust optical neural network design device against incident signal errors, comprising:
the acquisition module is used for acquiring an incident signal error caused by the loading and conversion process of the electrical input signal;
the first determining module is used for determining random noise with a target size range and probability distribution according to the incident signal error, and superposing the random noise on an original optical input signal to obtain an optical input signal based on the random noise;
and the training module is used for generating a training set based on the optical input signal based on the random noise, and training the optical neural network by using the training set until the training is finished to obtain the robust optical neural network with the function of resisting the incident signal error.
6. The apparatus of claim 5, wherein the random noise is a combined model of one or more physical properties of light for eliminating errors due to signal transitions and loading during computational reasoning of the optical neural network chip.
7. The apparatus as recited in claim 5, further comprising:
the second determining module is used for determining a network weight parameter according to the robust optical neural network with the anti-incident signal error after the robust optical neural network with the anti-incident signal error is obtained;
and determining processing parameters of the optical neural network chip based on the network weight parameters, and processing the optical neural network chip by utilizing the processing parameters.
8. The apparatus of any of claims 5-7, wherein the optical neural network comprises any one of a diffractive neural network, an interfering neural network, and a scattering neural network.
9. An electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the robust optical neural network design method against incident signal errors of any of claims 1-4.
10. A computer-readable storage medium having stored thereon a computer program, wherein the program is executed by a processor for implementing the robust optical neural network design method against incident signal errors of any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310325450.7A CN116542305A (en) | 2023-03-29 | 2023-03-29 | Robust optical neural network design method and device for resisting incident signal errors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310325450.7A CN116542305A (en) | 2023-03-29 | 2023-03-29 | Robust optical neural network design method and device for resisting incident signal errors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116542305A true CN116542305A (en) | 2023-08-04 |
Family
ID=87451328
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310325450.7A Pending CN116542305A (en) | 2023-03-29 | 2023-03-29 | Robust optical neural network design method and device for resisting incident signal errors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116542305A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140186057A1 (en) * | 2011-06-29 | 2014-07-03 | Alcatel Lucent | Method of demodulating a phase modulated optical signal |
| CN106227701A (en) * | 2016-06-30 | 2016-12-14 | 电子科技大学 | A kind of automatic correcting method of the amplitude phase error receiving passage of array signal |
| CN110728230A (en) * | 2019-10-10 | 2020-01-24 | 江南大学 | Signal modulation mode identification method based on convolution limited Boltzmann machine |
| US20220019883A1 (en) * | 2020-07-20 | 2022-01-20 | Nxp B.V. | Adc compensation using machine learning system |
| CN114418082A (en) * | 2022-01-24 | 2022-04-29 | 清华大学 | Parameter generation method and manufacturing method of optical neural network chip |
| CN115290125A (en) * | 2022-10-10 | 2022-11-04 | 泉州昆泰芯微电子科技有限公司 | Method for signal trimming by injecting random noise and magnetic encoder |
| CN115642970A (en) * | 2022-09-16 | 2023-01-24 | 北京交通大学 | Self-learning channel modeling method and system |
-
2023
- 2023-03-29 CN CN202310325450.7A patent/CN116542305A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140186057A1 (en) * | 2011-06-29 | 2014-07-03 | Alcatel Lucent | Method of demodulating a phase modulated optical signal |
| CN106227701A (en) * | 2016-06-30 | 2016-12-14 | 电子科技大学 | A kind of automatic correcting method of the amplitude phase error receiving passage of array signal |
| CN110728230A (en) * | 2019-10-10 | 2020-01-24 | 江南大学 | Signal modulation mode identification method based on convolution limited Boltzmann machine |
| US20220019883A1 (en) * | 2020-07-20 | 2022-01-20 | Nxp B.V. | Adc compensation using machine learning system |
| CN113965198A (en) * | 2020-07-20 | 2022-01-21 | 恩智浦有限公司 | ADC compensation using machine learning system |
| CN114418082A (en) * | 2022-01-24 | 2022-04-29 | 清华大学 | Parameter generation method and manufacturing method of optical neural network chip |
| CN115642970A (en) * | 2022-09-16 | 2023-01-24 | 北京交通大学 | Self-learning channel modeling method and system |
| CN115290125A (en) * | 2022-10-10 | 2022-11-04 | 泉州昆泰芯微电子科技有限公司 | Method for signal trimming by injecting random noise and magnetic encoder |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI591490B (en) | Vector computation unit in a neural network processor | |
| KR102215271B1 (en) | Model calculation unit, control device and method for calculating a data-based function model | |
| US12014130B2 (en) | System and method for ESL modeling of machine learning | |
| KR102448018B1 (en) | Method and apparatus for testing memory, electronic device, storage medium and program | |
| EP3649582A1 (en) | System and method for automatic building of learning machines using learning machines | |
| CN116523015A (en) | Optical neural network training method, device and equipment for process error robustness | |
| CN114548027A (en) | Method for tracking signal in verification system, electronic device and storage medium | |
| CN111797588B (en) | Formal verification comparison point matching method, system, processor and memory | |
| US20210287077A1 (en) | Systems and methods for implementing operational transformations for restricted computations of a mixed-signal integrated circuit | |
| CN116663491B (en) | Method, equipment and medium for covering group condition constraint statement based on BDD solving function | |
| EP3745319A1 (en) | Optimization apparatus and optimization method | |
| CN116542305A (en) | Robust optical neural network design method and device for resisting incident signal errors | |
| KR102255470B1 (en) | Method and apparatus for artificial neural network | |
| CN116663493A (en) | Conditional constraint statement solving method, device and medium based on constraint solver | |
| CN115809707B (en) | Quantum comparison operation method, device, electronic device and basic arithmetic component | |
| CN116467877A (en) | Floating wind turbine generator system platform dynamic response determination method and device and electronic equipment | |
| CN116384460B (en) | Robust optical neural network training method and device, electronic equipment and medium | |
| CN102063308B (en) | Method for controlling processing flow of seismic prospecting data | |
| CN113312862B (en) | LFSR-based random circuit hardware overhead minimization design method | |
| CN113760751B (en) | Method for generating test case, electronic device and storage medium | |
| CN109933948B (en) | Form verification method, device, form verification platform and readable storage medium | |
| Guo et al. | An orchestrated empirical study on deep learning frameworks and platforms | |
| JP2022124240A (en) | Diagnostic pattern generation method and computer | |
| Reimer et al. | Using maxbmc for pareto-optimal circuit initialization | |
| KR102412872B1 (en) | Processing element, method of operation thereof, and accelerator including the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |