WO2020048375A1 - 量子比特检测系统及检测方法 - Google Patents

量子比特检测系统及检测方法 Download PDF

Info

Publication number
WO2020048375A1
WO2020048375A1 PCT/CN2019/103246 CN2019103246W WO2020048375A1 WO 2020048375 A1 WO2020048375 A1 WO 2020048375A1 CN 2019103246 W CN2019103246 W CN 2019103246W WO 2020048375 A1 WO2020048375 A1 WO 2020048375A1
Authority
WO
WIPO (PCT)
Prior art keywords
qubit
module
machine learning
learning model
imaging device
Prior art date
Application number
PCT/CN2019/103246
Other languages
English (en)
French (fr)
Inventor
徐华
Original Assignee
阿里巴巴集团控股有限公司
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 阿里巴巴集团控股有限公司 filed Critical 阿里巴巴集团控股有限公司
Priority to JP2021510187A priority Critical patent/JP2021535483A/ja
Priority to EP19858129.0A priority patent/EP3848860A4/en
Publication of WO2020048375A1 publication Critical patent/WO2020048375A1/zh
Priority to US17/189,719 priority patent/US20210182727A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N10/00—Quantum computing, i.e. information processing based on quantum-mechanical phenomena
    • G06N10/40—Physical realisations or architectures of quantum processors or components for manipulating qubits, e.g. qubit coupling or qubit control
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J4/00—Measuring polarisation of light
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N10/00—Quantum computing, i.e. information processing based on quantum-mechanical phenomena
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • 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

Definitions

  • the present disclosure relates to the field of detection, and in particular, to a qubit detection system and a detection method.
  • Quantum computing and quantum information are an interdisciplinary discipline based on the principles of quantum mechanics to achieve computing and information processing tasks. They are closely related to quantum physics, computer science, and information science. It has developed rapidly in the last two decades. Quantum computer-based quantum algorithms in scenarios such as factorization and unstructured search show performance that far surpasses existing classical computer-based algorithms, and this direction is also expected to exceed existing computing capabilities.
  • the main implementation methods of qubits include: superconducting Josephson junctions, ion traps, magnetic resonance, topological qubits, etc.
  • researchers generally use electronic equipment to test the prepared qubits to determine whether the prepared qubits are Defective and get its preliminary properties parameters. This method is time-consuming and labor-intensive, especially for qubits based on the superconducting Josephson junction, which needs to be detected in a low temperature environment (liquid helium temperature), and the cost of obtaining a low temperature environment is very high. The cost of testing in low temperature environments is also very high.
  • the detection of qubits in the prior art can only be performed after the qubit preparation is completed, that is, if a defect occurs in the preparation process of the qubit, the detection methods based on the prior art cannot be known .
  • the prior art's detection methods for qubits are not only costly, but also difficult to meet future needs after the large-scale industrialization of the preparation of qubits.
  • a qubit detection method including: imaging an qubit using an imaging device; inputting the obtained image into a machine learning model; and the machine learning model outputs prediction information based at least on the obtained image.
  • a detection device is used to detect the qubit to obtain detection data; the detection data is input to a machine learning model; the machine learning model outputs prediction information based on the test data and the obtained image.
  • the imaging device performs imaging during the preparation process of the qubit, and the preparation process includes multiple steps.
  • an imaging device is used to perform imaging after the qubit preparation is completed.
  • the prediction information includes quantitative information and qualitative information.
  • the quantitative information includes at least one of the following: operating frequency, coherence time, and coupling strength.
  • the qualitative information includes: a gradation of a qubit.
  • a qubit detection system including: a test module, the test module includes an imaging device, the imaging device is configured to image the qubit, and the prediction module, the prediction module, and the test module are communicatively connected
  • the prediction module includes a machine learning model, and the machine learning model is configured to output prediction information based at least on the obtained image.
  • the test module further includes a detection device configured to detect the qubit to obtain detection data.
  • the imaging device includes: a scanning electron microscope (SEM) and a scanning tunnel microscope (STM).
  • SEM scanning electron microscope
  • STM scanning tunnel microscope
  • the machine learning model includes: a convolutional neural network (CNN), a deep neural network (DNN), and a regression neural network (RNN).
  • CNN convolutional neural network
  • DNN deep neural network
  • RNN regression neural network
  • the qubit detection system further includes: an interaction module and a monitoring module, a communication connection between the monitoring module and the prediction module, and a communication connection between the interaction module and the prediction module and the monitoring module.
  • the qubit detection system further includes: a test control module, the test control module is located between the test module and the prediction module, the test control module is in communication with the test module and the prediction module, and the test control module includes: imaging data generation Module, imaging data reading module, qubit test generation module, qubit test data reading module.
  • the prediction module includes a qubit defect determination and property prediction module.
  • a qubit detection device including: a processor and a non-transitory storage medium.
  • the non-transitory storage medium stores an instruction set, and the instruction set is executed by a processor. Realized at time: means for causing an imaging device to image a qubit; means for inputting the obtained image into a machine learning model; means for causing a machine learning model to output prediction information based at least on the obtained image.
  • 1 is a block diagram of a qubit detection system based on some embodiments of the present invention.
  • 100 qubit detection system
  • 200 qubit preparation process
  • 2 imaging device
  • 22 detection device
  • 32 imaging data reading Modules
  • 33 qubit test generation module
  • 34 qubit test data reading module
  • 42 machine learning model
  • the functional blocks of some embodiments do not necessarily indicate division between hardware circuits.
  • one or more of the functional blocks may be implemented in a single piece of hardware (such as a general-purpose signal processor or a piece of random access memory, a hard disk, etc.) or multiple pieces of hardware.
  • the program may be an independent program, which may be combined into a routine in an operating system, or a function in an installed software package, and the like. It should be understood that some embodiments are not limited to the arrangements and tools shown in the figures.
  • FIG. 1 shows a qubit detection system 100 according to some embodiments.
  • the qubit system 100 shown on the right side of FIG. 1 includes a test module 2, a test control module 3, a prediction module 4, an interaction module 5 and a monitoring module 6.
  • the left side of FIG. 1 shows a qubit preparation process 200.
  • the process 200 may include multiple steps, as shown in the figure: step 1, step 2 ... step k ... step n.
  • the test module 2 includes an imaging device 21 and a detection device 22.
  • the imaging device 21 may be a scanning electron microscope (SEM), and the detection device 22 may be any conventional device that can detect qubits. Device.
  • the scanning electron microscope 21 can be used to image qubits, and imaging can occur at any step in the process of preparing qubits.
  • the imaging device 22 may further include other imaging devices known to those skilled in the art, such as infrared imaging devices, optical imaging devices in the visible light frequency range, and the like. Using these devices can process Perform imaging in any one or more of the steps.
  • the test control module 3 includes an imaging data generating module 31, an imaging data reading module 32, a qubit test generating module 33, and a qubit test data reading module 34.
  • the test control module 3 and the test module 2 are communicatively connected.
  • the imaging data generating module 31 is used to generate test requirements (such as specific imaging parameters, imaging time, etc.) required for imaging, and send the generated imaging test requirements to the imaging device 21, so that the imaging device 21 needs to perform the required tests according to the test requirements. Test to obtain the formed image and the parameters collected during imaging. Then, the formed image and the parameters collected during imaging are transmitted to the imaging data reading module 32.
  • the imaging device 21 is a scanning electron microscope (SEM)
  • the data generation module 31 when the imaging device 21 is a scanning electron microscope (SEM), the data generation module 31 generates a SEM test request, and sends the test request to the SEM, and controls the SEM to perform the required test and generate the SEM image and the SEM image.
  • the parameters at the time of acquisition include at least one image, usually multiple images. Multiple images can be acquired continuously or at intervals. These acquired SEM images and parameters during SEM image acquisition are sent to the imaging data reading module 32.
  • the data generation module 31 when the imaging device 21 is a scanning tunneling microscope (STM), the data generation module 31 generates an STM test request and sends the test request to the STM to control the STM to perform the required tests and generate STM images and STM images
  • the parameters at the time of acquisition include at least one image, usually multiple images. Multiple images can be acquired continuously or at intervals. These acquired STM images and parameters during STM image acquisition are sent to the imaging data reading module 32.
  • the imaging device 21 is an optical imaging device in the visible light frequency range, such as a photosensitive coupling device (CCD).
  • the optical imaging device is used to image one or more steps of the qubit preparation process, and the acquired image is at least One image is usually included, and multiple images can be acquired continuously or at intervals. These acquired images are sent to the imaging data reading module 32.
  • the imaging device 21 is an infrared imaging device.
  • the infrared imaging device is used to image one or more steps of the qubit preparation process.
  • the acquired image includes at least one, and usually includes multiple images. Continuous acquisition or interval acquisition. These acquired images are sent to the imaging data reading module 32.
  • the imaging device 21 is an ultraviolet imaging device.
  • the ultraviolet imaging device is used to image one or more steps of the qubit preparation process.
  • the acquired image includes at least one, and usually includes multiple images. Continuous acquisition or interval acquisition. These acquired images are sent to the imaging data reading module 32.
  • the qubit test generation module 33 is configured to generate test requirements required for detecting qubits, and send the generated test requirements to the detection device 22, so that the detection device 22 performs required tests according to the test requirements and obtains test results. These test results may include: the operating frequency of the qubit, the coherence time of the qubit, the coupling strength of the qubit, and so on. The obtained test result is sent to the qubit test data reading module 34.
  • the prediction module 4 includes a qubit defect determination and property prediction module 41 and a machine learning model 42.
  • the prediction module 4 and the test control module 3 are communicatively connected.
  • the machine learning model 42 may be a trained convolutional neural network (CNN).
  • the training data of the convolutional neural network may include: a qubit design drawing (gds file) input by the user, an image (including the image itself, and image acquisition) obtained in the qubit preparation process transmitted via the imaging data reading module 33 Parameters, test requirements required for imaging, etc.); and test results obtained by testing the qubit through the detection device 22 transmitted by the qubit test data reading module 34, and test parameters. These test results may include : The operating frequency of the qubit, the coherence time of the qubit, the coupling strength of the qubit, etc.
  • the imaging device is a SEM
  • the training data of the convolutional neural network includes: a qubit design drawing (gds file) input by the user, obtained in a qubit preparation process transmitted through the imaging data reading module 33 SEM image (including the SEM image itself, parameters during SEM image acquisition, test requirements required for SEM imaging, etc.); and obtained by testing the qubit through the detection device 22 transmitted by the qubit test data reading module 34 Test results, and test parameters.
  • Test results can include: the operating frequency of the qubit, the coherence time of the qubit, the coupling strength of the qubit, and so on.
  • the imaging device is an STM
  • the training data of the convolutional neural network includes: a qubit design drawing (eg, but not limited to: a gds file) input by the user, and a qubit transmitted via the imaging data reading module 33
  • the STM image obtained during the preparation process (including the STM image itself, as well as parameters during STM image acquisition, test requirements required for STM imaging, etc.); and the detection device 22 transmitted by the qubit test data reading module 34 to the quantum
  • These test results may include: the operating frequency of the qubit, the coherence time of the qubit, the coupling strength of the qubit, and so on.
  • the imaging device is an optical device.
  • the optical device includes an imaging device in a visible light frequency range, or an infrared imaging device, similar to the above SEM or STM imaging device. These imaging devices in a visible light frequency range And the image called by the infrared imaging device, the parameters during imaging, and the test requirements required for imaging can all be used as training inputs for the convolutional neural network.
  • the properties of the qubit can be determined and predicted based on the actual input, and the results of the determination and prediction are transmitted to the qubit defect determination and property prediction module 41.
  • the actual inputs include the results obtained from the tests performed during the qubit preparation process. Specifically, for example, using a SEM device to image the qubits during the qubit preparation process.
  • the imaging can occur in any one of the preparation processes Step, or multiple steps, the imaging data is transmitted to the trained convolutional neural network and the quantum information is given by the convolutional neural network, including the operating frequency of the qubit and the coherence time of the qubit. Coupling strength of qubits, etc.
  • These quantitative information are further transmitted to the qubit defect determination and property prediction module for output.
  • the qubit defect determination and property prediction module can directly output quantitative information, or can give qualitative information based on the quantitative information, such as grading of qubit properties.
  • the required training of the convolutional neural network may be performed according to different types of qubits and different qubit designs.
  • the machine learning model 42 may be a deep neural network (DNN) or a regression neural network (RNN).
  • DNN deep neural network
  • RNN regression neural network
  • the training process and output process of DNN and RNN are similar to the convolutional neural network (CNN) described above.
  • CNN convolutional neural network
  • Those skilled in the art may also use any other algorithm known in the art as the machine learning model 42.
  • the qubit detection system 100 further includes an interaction module 5 and a monitoring module 6, the interaction module 5 and the prediction module 4 and the monitoring module 6 are communicatively connected, the monitoring module 6 and the test control module 3, the prediction module 4, and the interaction module 5 and a qubit preparation flow 100 communication connection.
  • the interaction module 5 is used to support the input and output functions of the user, transmit the user's input information and various parameters to the prediction module 4 and the monitoring module 6 as required, and display the prediction result output by the prediction module 4 to the user.
  • the monitoring module 6 monitors the test control module 3, the prediction module 4, the interaction module 5, and the qubit preparation process 100.
  • the predicted qualitative and quantitative information is further transmitted to the interaction module 5 and the monitoring module 6, and the user obtains the predicted qualitative information through the monitoring module 6.
  • the quantitative and quantitative information based on these qualitative and quantitative information, users can instantly adjust various parameters in the qubit preparation process.
  • These adaptive adjustments include, but are not limited to, a single preparation step (such as step n ) To make adjustments, and the entire qubit preparation process can be optimized as a whole after the qubit preparation is completed. Thus, the yield of the qubit can be improved (Yield).
  • a qubit detection method includes: imaging an qubit using an imaging device; inputting the obtained image into a trained machine learning model.
  • the training process may be as described above, and the machine learning model outputs prediction information based on at least the obtained image .
  • the imaging device is a SEM and the machine learning model is a convolutional neural network (CNN).
  • CNN convolutional neural network
  • the imaging device is an STM and the machine learning model is a convolutional neural network (CNN).
  • CNN convolutional neural network
  • the imaging device is an imaging device in the visible light frequency range
  • the machine learning model is a convolutional neural network (CNN).
  • the imaging device is an infrared imaging device and the machine learning model is a convolutional neural network (CNN).
  • CNN convolutional neural network
  • the imaging device is a UV imaging device and the machine learning model is a convolutional neural network (CNN).
  • CNN convolutional neural network
  • the machine learning model is a deep neural network (DNN).
  • DNN deep neural network
  • the machine learning model is a regression neural network (RNN).
  • RNN regression neural network
  • one or more processes are selected in the process of the qubit preparation device to use SEM for imaging, and the obtained image is input into a trained convolutional neural network model, and the convolutional neural network model is based on the obtained image Output prediction information.
  • the prediction information includes quantitative information, such as the operating frequency of the qubit, the coherence time of the qubit, the coupling strength of the qubit, and so on. It can also include qualitative information, such as the nature classification of the qubit. Based on these prediction information, the user can quickly know the properties of the qubit at a certain step or several steps, and adjust the qubit preparation process 100 accordingly based on these properties.
  • one or more processes are selected in the process of the qubit preparation device to use STM for imaging, and the obtained image is input to a trained convolutional neural network model, and the convolutional neural network model is based on the obtained image Output prediction information.
  • the prediction information includes quantitative information, such as the operating frequency of the qubit, the coherence time of the qubit, the coupling strength of the qubit, and so on. It can also include qualitative information, such as the nature classification of the qubit. Based on these prediction information, the user can quickly know the properties of the qubit at a certain step or several steps, and adjust the qubit preparation process 100 accordingly based on these properties.
  • the system and method of the present disclosure can be effectively integrated with existing and possible large-scale production lines of qubits, and a quick estimation of the properties of the qubits is made during the qubit production process or after the qubit production preparation is completed. Fast estimation, can perform more efficient qubit detection.
  • the necessary parameter adjustment and optimization of the qubit production process can be efficiently performed, thereby improving the overall yield of the qubit production line (Yield), and the parameter adjustment and optimization of the present disclosure It is not limited to the completion of the qubit preparation, but can be adjusted immediately during the qubit preparation process, which can be more conducive to improving the yield in the qubit production process.
  • the present disclosure uses an imaging device to detect qubits and analyzes the images obtained by the quanta according to a machine learning model, thereby obtaining qualitative and quantitative information predictions of the qubits.
  • the technical solution of the present disclosure The cost is extremely low, and it has great industrial application advantages after the mass production of qubits is realized.
  • Computer-readable media includes both permanent and non-persistent, removable and non-removable media.
  • Information can be stored by any method or technology.
  • Information may be computer-readable instructions, data structures, modules of a program, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), and read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, read-only disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media may be used to store information that can be accessed by computing devices.
  • computer-readable media does not include temporary computer-readable media, such as modulated data signals and carrier waves.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Artificial Intelligence (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computational Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Image Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Holo Graphy (AREA)

Abstract

一种量子比特检测系统(100)及方法，检测系统(100)包括测试模块(2)和预测模块(4)，测试模块(2)包括成像装置(21)，成像装置(21)配置为对量子比特进行成像；预测模块(4)和测试模块(2)通信连接，预测模块(4)包括机器学习模型(42)，机器学习模型(42)配置为至少基于得到的像输出预测信息，检测方法包括：使用成像装置(21)对量子比特进行成像；将得到的像输入训练好的机器学习模型(42)；机器学习模型(42)至少基于得到的像输出预测信息。

Description

量子比特检测系统及检测方法
本申请要求2018年09月05日递交的申请号为201811029930.4、发明名称为“量子比特检测系统及检测方法”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本披露涉及检测领域，特别地，涉及一种量子比特的检测系统以及检测方法。
背景技术
量子计算与量子信息是一门基于量子力学的原理来实现计算与信息处理任务的交叉学科，与量子物理，计算机科学，信息学等学科有着十分紧密的联系。在最近二十年有着快速的发展。因数分解，无结构搜索等场景的基于量子计算机的量子算法展现出了远超越现有基于经典计算机的算法的表现，也使这一方向被寄予了超越现有计算能力的期望。
量子计算机的基本特点之一就是，它使用的信息单元不是比特，而是量子比特(Qubit)。量子比特可以是电子那样的粒子，也可以是其它的处于元激发的准粒子。对于电子而言，自旋向上代表1，向下代表0。自旋既向上又向下的量子态称为叠加态。处于叠加态中的少量粒子可以携带大量信息，仅仅100个粒子所处的叠加态，就可以表示从1到2100个数字。量子计算机可以用激光脉冲打击粒子，或者采用诸如此类的方法来对量子比特进行操作。
目前，量子比特的主要实现方式包括：超导约瑟夫森结、离子阱、磁共振、拓扑量子等；研究人员一般使用电子设备仪器对制备得到的量子比特进行检测，以确定制备得到的量子比特是否有缺陷，并获得其初步的性质参数。这种方法耗时耗力，尤其是对于基于超导约瑟夫森结的量子比特，对其进行检测需要在低温环境(液氦温度)下进行，而低温环境的获得需要的费用很高，因此在低温环境下进行检测所需要的费用也非常高。另一方面，现有技术对于量子比特的检测只能在量子比特制备完成后进行，也即是说，如果量子比特的制备过程中出现了缺陷，基于现有技术的检测手段是无法得知的。基于以上，现有技术对于量子比特的检测手段不仅耗费甚巨，并且也难以满足未来当量子比特的制备大规模工业化之后的需求。
基于以上，需要一种量子比特的检测装置以及方法，以解决上述的这些技术问题。
发明内容
根据本披露的一个方面的实施例，提供一种量子比特检测方法，包括：使用成像装置对量子比特进行成像；将得到的像输入机器学习模型；机器学习模型至少基于得到的像输出预测信息。
作为优选，使用检测装置对量子比特进行检测，得到检测数据；将检测数据输入机器学习模型；机器学习模型基于测试数据和得到的像输出预测信息。
作为优选，获取量子比特的设计信息；将量子比特的设计信息输入机器学习模型；机器学习模型基于设计信息、测试数据和得到的像输出预测信息。
作为优选，成像装置包括扫描电子显微镜(SEM)和扫描隧道显微镜(STM)。
作为优选，成像装置在量子比特的制备过程中进行成像，制备过程包括多个步骤。
作为优选，使用成像装置在量子比特制备完成后进行成像。
作为优选，预测信息包括定量信息和定性信息。
作为优选，定量信息包括以下至少一种：工作频率、相干时间、耦合强度。
作为优选，定性信息包括：量子比特的分级。
根据本发明另一方面的一些实施例，提供一种量子比特检测系统，包括：测试模块，测试模块包括成像装置，成像装置配置为对量子比特进行成像；预测模块，预测模块和测试模块通信连接，预测模块包括机器学习模型，机器学习模型配置为至少基于得到的像输出预测信息。
作为优选，测试模块还包括检测装置，检测装置配置为对量子比特进行检测以得到检测数据。
作为优选，成像装置包括：扫描电子显微镜(SEM)和扫描隧道显微镜(STM)。
作为优选，机器学习模型包括：卷积神经网络(CNN)、深度神经网络(DNN)、回归神经网络(RNN)。
作为优选，量子比特检测系统还包括：交互模块和监控模块，监控模块和预测模块通信连接，交互模块和预测模块以及监控模块通信连接。
作为优选，量子比特检测系统还包括：测试控制模块，测试控制模块位于所述测试模块和所述预测模块之间，测试控制模块和测试模块以及预测模块通信连接，测试控制模块包括：成像数据产生模块、成像数据读取模块、量子比特测试产生模块、量子比特 测试数据读取模块。
作为优选，预测模块包括：量子比特缺陷判定和性质预测模块。
根据本发明另一方面的一些实施例，提供一种量子比特检测装置，包括：处理器和非暂态存储介质，所述非暂态存储介质存储有指令集，所述指令集被处理器执行时实现：用于使成像装置对量子比特进行成像的装置；用于将所述得到的像输入机器学习模型的装置；用于使机器学习模型至少基于所述得到的像输出预测信息的装置。
附图说明
此处所说明的附图用来提供对本披露的进一步理解，构成本披露的一部分，本披露的示意性实施例及其说明用于解释本披露，并不构成对本披露的不当限定。在附图中：
图1为基于本发明一些实施例的量子比特检测系统的框架图；
图2为基于本发明一些实施例的量子比特检测方法的流程图；
其中：100：量子比特检测系统、200：量子比特制备流程、2、测试模块；21：成像装置；22：检测装置；3、测试控制模块；31、成像数据产生模块；32、成像数据读取模块；33、量子比特测试产生模块；34、量子比特测试数据读取模块；4、预测模块；41、量子比特缺陷判定和性质预测模块；42、机器学习模型；5、交互模块；6、监控模块。
具体实施方式
当结合附图来阅读时，将更好地理解前述概述以及某些实施例的以下详细描述。就图示出一些实施例的功能框的简图而言，功能框未必指示硬件电路之间的分割。因而，例如，可在单件硬件(例如通用信号处理器或一块随机存取存储器、硬盘等)或多件硬件中实施功能框中的一个或多个(例如处理器或存储器)。类似地，程序可为独立的程序，可结合成操作系统中的例程，可为安装好的软件包中的函数等。应当理解，一些实施例不限于图中显示的布置和工具。
如本披露所用，以单数叙述或以词语“一个”或“一种”开头的要素或步骤应理解为不排除所述要素或步骤的复数，除非明确陈述了这种排除。此外，对“一个实施例”的引用不意于被解释为排除也结合了所叙述的特征的另外的实施例的存在。除非明确陈述了相反的情况，否则“包括”、“包含”或“具有”具有特定属性的要素或多个要素的实施例可包括不具有那个属性的另外的这样的要素。
图1示出了根据一些实施例的量子比特检测系统100，图1右侧示出的量子比特系 统100包括测试模块2，测试控制模块3，预测模块4，交互模块5和监控模块6。图1左侧示出了量子比特制备流程200，流程200可以包括多个步骤，如图所示意的：步骤1、步骤2…步骤k…步骤n。
在一些实施例中，测试模块2包括成像装置21和检测装置22，成像装置21可以是扫描电子显微镜(SEM)，检测装置22可以是现有技术中任何习知的可以对量子比特进行检测的装置。扫描电子显微镜21可以用于对量子比特进行成像，成像可以发生在量子比特制备过程中的任意一个步骤。
在一些实施例中，成像装置22还可以是扫描隧道显微镜(STM)，而检测装置22可以是现有技术中任何习知的可以对量子比特进行检测的电子装置。扫描隧道显微镜22可以用于对量子比特进行成像，成像可以发生在量子比特制备过程中的任意一个步骤。
在一些实施例中，成像装置22还可以包括其它的本领域技术人员所习知的成像装置，例如红外成像装置，可见光频率范围内的光学成像装置等，使用这些装置可以对量子比特在制备过程的任意一个或多个步骤进行成像。
在一些实施例中，测试控制模块3包括：成像数据产生模块31、成像数据读取模块32、量子比特测试产生模块33以及量子比特测试数据读取模块34。测试控制模块3和测试模块2通信连接。
成像数据产生模块31用于产生成像所需的测试要求(例如具体的成像参数、成像时间等)，并将产生的成像测试要求发送给成像装置21，从而使得成像装置21按照测试要求进行所需要的测试，获取所成的像以及成像时所采集的参数。然后，所成的像和成像时所采集的参数被发送给成像数据读取模块32。在一些实施例中，当成像装置21是扫描电子显微镜(SEM)时，数据产生模块31产生SEM测试要求，并将测试要求发送给SEM，控制SEM进行所需要的测试并产生SEM图像以及SEM图像采集时的参数，所获取的图像至少包括一幅，通常包括多幅，多幅图像可以连续获取，也可以间隔时间获取。这些获取到的SEM图像以及SEM图像采集时的参数被发送给成像数据读取模块32。
在一些实施例中，当成像装置21是扫描隧道显微镜(STM)时，数据产生模块31产生STM测试要求，并将测试要求发送给STM，控制STM进行所需要的测试并产生STM图像以及STM图像采集时的参数，所获取的图像至少包括一幅，通常包括多幅，多幅图像可以连续获取，也可以间隔时间获取。这些获取到的STM图像以及STM图像采集时的参数被发送给成像数据读取模块32。
在一些实施例中，成像装置21是可见光频率范围内的光学成像装置，例如感光耦合装置(CCD)，使用光学成像装置对量子比特制备过程的一个或多个步骤进行成像，所获取的图像至少包括一幅，通常包括多幅，多幅图像可以连续获取，也可以间隔时间获取。这些获取到的图像被发送给成像数据读取模块32。
在一些实施例中，成像装置21是红外成像装置，使用红外成像装置对量子比特制备过程的一个或多个步骤进行成像，所获取的图像至少包括一幅，通常包括多幅，多幅图像可以连续获取，也可以间隔时间获取。这些获取到的图像被发送给成像数据读取模块32。
在一些实施例中，成像装置21是紫外成像装置，使用紫外成像装置对量子比特制备过程的一个或多个步骤进行成像，所获取的图像至少包括一幅，通常包括多幅，多幅图像可以连续获取，也可以间隔时间获取。这些获取到的图像被发送给成像数据读取模块32。
量子比特测试产生模块33用于产生检测量子比特所需要的测试要求，并将产生的测试要求发送给检测装置22，从而使得检测装置22按照测试要求进行所需要的测试，并获取测试结果。这些测试结果可以包括：量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度等。获取到的测试结果被发送给量子比特测试数据读取模块34。
在一些实施例中，预测模块4包括量子比特缺陷判定和性质预测模块41和机器学习模型42。预测模块4和测试控制模块3通信连接。
在一些实施例中，机器学习模型42可以是训练好的卷积神经网络(CNN)。卷积神经网络的训练数据可以包括：由用户输入的量子比特设计图纸(gds文件)，经由成像数据读取模块33所传输的量子比特制备流程中所获取的图像(包括图像本身，以及图像采集时的参数，成像所需的测试要求等)；以及经由量子比特测试数据读取模块34所传输的检测装置22对量子比特进行测试所获得的测试结果，以及测试参数等，这些测试结果可以包括：量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度等。
在一些实施例中，成像装置为SEM，卷积神经网络的训练数据包括：由用户输入的量子比特设计图纸(gds文件)，经由成像数据读取模块33所传输的量子比特制备流程中所获取的SEM图像(包括SEM图像本身，以及SEM图像采集时的参数，SEM成像所需的测试要求等)；以及经由量子比特测试数据读取模块34所传输的检测装置22对 量子比特进行测试所获得的测试结果，以及测试参数等，这些测试结果可以包括：量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度等。
在一些实施例中，成像装置为STM，卷积神经网络的训练数据包括：由用户输入的量子比特设计图纸(例如但不限于：gds文件)，经由成像数据读取模块33所传输的量子比特制备流程中所获取的STM图像(包括STM图像本身，以及STM图像采集时的参数，STM成像所需的测试要求等)；以及经由量子比特测试数据读取模块34所传输的检测装置22对量子比特进行测试所获得的测试结果，以及测试参数等，这些测试结果可以包括：量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度等。
在一些实施例中，成像装置为光学装置，特别地，光学装置包括可见光频率范围内的成像装置，或者红外成像装置，与上面的SEM或STM的成像装置类似，这些可见光频率范围内的成像装置以及红外成像装置所称的像，成像时的参数，成像所需的测试要求等都可以作为卷积神经网络的训练输入。
在一些实施例中，当卷积神经网络训练完成后，即可基于实际的输入来对量子比特的性质进行判定和预测，判定和预测的结果传输至量子比特缺陷判定和性质预测模块41。实际的输入包括在量子比特制备流程中所进行的测试所得到的结果，具体而言，使用例如SEM装置在量子比特制备的流程中对量子比特进行SEM成像，成像可以发生在制备流程的任意一个步骤，或者多个步骤，成像数据被传输至训练好的卷积神经网络并由卷积神经网络给出量子比特的定量信息，包括量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度等。这些定量信息被进一步传输至量子比特缺陷判定和性质预测模块进行输出，量子比特缺陷判定和性质预测模块可以直接输出定量信息，也可以基于定量信息给出定性信息，例如量子比特性质的分级。
在一些实施例中，可以依据不同类型的量子比特，以及不同的量子比特设计来对卷积神经网络进行所需要的训练。另外，还可以定期地抽样制备的量子比特样品进行测试，获取量子比特的性质测试值(例如但不限于量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度)，通过比较真实的量子比特性质测量值与卷积神经网络的预测值，从而决定是否需要对卷积神经网络进行重新训练调节。
在另一些实施例中，机器学习模型42可以是深度神经网络(DNN)，或者是回归神经网络(RNN)，DNN和RNN的训练过程和输出过程和上述的卷积神经网络(CNN)相类似。本领域技术人员还可以采用其它任意本领域所习知的算法来作为机器学习模型 42。
在一些实施例中，量子比特检测系统100还包括交互模块5和监控模块6，交互模块5和预测模块4以及监控模块6通信连接，监控模块6和测试控制模块3、预测模块4、交互模块5以及量子比特制备流程100通信连接。交互模块5用于支持用户输入输出功能，将用户的输入信息以及各类参数按需要传输给预测模块4、监控模块6，并将预测模块4产生的预测结果输出展现给用户。监控模块6对测试控制模块3、预测模块4、交互模块5以及量子比特制备流程100进行监控。
在一些实施例中，当预测模块4获得了预测的定性和定量信息后，这些预测的定性和定量信息被进一步传输到交互模块5和监控模块6，用户通过监控模块6获得了这些预测的定性和定量信息后，基于这些定性和定量信息，用户可以即时地对量子比特制备流程中的各种参数进行适应性调整，这些适应性调整包括但不限于对单独的某个制备步骤(例如步骤n)进行调整，也可以在量子比特制备结束后对整个量子比特制备流程进行整体性优化。从而可以提升量子比特的制备良率(Yield)。
在一些实施例中，提供了一种量子比特的检测方法。如图2所示意，检测方法包括：使用成像装置对量子比特进行成像；将得到的像输入训练好的机器学习模型，训练过程可以如上文所述，机器学习模型至少基于得到的像输出预测信息。
在一些实施例中，成像装置是SEM，机器学习模型是卷积神经网络(CNN)。
在一些实施例中，成像装置是STM，机器学习模型是卷积神经网络(CNN)。
在一些实施例中，成像装置是可见光频率范围内的成像装置，机器学习模型是卷积神经网络(CNN)。
在一些实施例中，成像装置是红外成像装置，机器学习模型是卷积神经网络(CNN)。
在一些实施例中，成像装置是紫外成像装置，机器学习模型是卷积神经网络(CNN)。
在一些实施例中，机器学习模型是深度神经网络(DNN)。
在一些实施例中，机器学习模型是回归神经网络(RNN)。
在一些实施例中，在量子比特制备装置流程中选择一个或者多个流程使用SEM进行成像，将所得到的像输入训练好的卷积神经网络模型，并由卷积神经网络模型基于得到的像输出预测信息。预测信息包括定量信息，例如量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度等，也可以包括定性信息，例如量子比特的性质分级。基于这些预测信息，用户可以快速地知道在某一个步骤或者在某几个步骤时量子比特的性质，以及基于这些性质相应地对量子比特制备流程100进行相应地调 整。
在一些实施例中，在量子比特制备装置流程中选择一个或者多个流程使用STM进行成像，将所得到的像输入训练好的卷积神经网络模型，并由卷积神经网络模型基于得到的像输出预测信息。预测信息包括定量信息，例如量子比特的工作频率，量子比特的相干时间(Coherence time)，量子比特的耦合强度等，也可以包括定性信息，例如量子比特的性质分级。基于这些预测信息，用户可以快速地知道在某一个步骤或者在某几个步骤时量子比特的性质，以及基于这些性质相应地对量子比特制备流程100进行相应地调整。
各种实施例体统了一种量子比特检测的系统及检测方法，本披露的优势包括但不限于：
本披露的系统和方法可以有效地与现有及未来可能的量子比特大规模生产工艺线集成，在量子比特生产进行过程中或者量子比特生产制备完成后进行量子比特性质的快速预估，根据这个快速预估，可以进行更高效率的量子比特检测。
基于本披露的系统和方法的快速预估，可以高效地对量子比特生产工艺流程进行必要的参数调整优化，从而可以提高量子比特生产线的整体良率(Yield)，并且，本披露的参数调整优化并不局限于在量子比特制备完成后才能进行，而是可以在量子比特的制备过程中即时地进行调整，从而可以更加有利于提升量子比特生产过程中的良率。
本披露采用成像装置对量子比特进行检测，并依据机器学习模型对量子所得到的像进行分析，从而得到量子比特的定性和定量信息预测，较之现有技术而言，本披露的技术方案的成本极低，在量子比特批量化大规模生产实现后具有极大的工业应用优势。
要理解的是，以上描述意于为示例性，而不是限制性的。例如，上面描述的实施例(和/或它们的各方面)可与彼此结合起来使用。另外，可在不偏离一些实施例的范围的情况下做出许多修改，以使具体情况或内容适于一些实施例的教导。虽然本文描述的材料的尺寸和类型意于限定一些实施例的参数，但实施例决不是限制性的，而是示例性实施例。在审阅以上描述之后，许多其它实施例对本领域技术人员将是显而易见的。因此，应当参照所附权利要求以及这样的权利要求所涵盖的等效体的全部范围来确定一些实施例的范围。在所附权利要求中，用语“包括”和“在其中”用作相应的用语“包含”和“其中”的易懂语言等效体。此外，在所附权利要求中，用语“第一”、“第二”和“第三”等仅作为标记使用，并且它们不意于对它们的对象施加数字要求。另外，不以手段加功能的格式来书写所附权利要求的限制，除非且直到这样的权利要求限制清楚地使用短语 “用于…的手段”，跟随没有另外的结构的功能陈述。
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。
本领域内的技术人员应明白，本披露的一些实施例可提供为方法、设备、或计算机程序产品。因此，本披露可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本披露可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。
本书面描述使用示例来公开一些实施例，包括最佳模式，并且还使本领域任何技术人员能够实践一些实施例，包括制造和使用任何装置或系统，以及实行任何结合的方法。一些实施例的保护范围由权利要求限定，并且可包括本领域技术人员想到的其它示例。如果这样的其它示例具有不异于权利要求的字面语言的结构要素，或者如果它们包括与权利要求的字面语言无实质性差异的等效结构要素，则它们意于处在权利要求的范围之内。

Claims (19)

  1. 一种量子比特检测方法，包括：
    使用成像装置对量子比特进行成像；
    将得到的像输入机器学习模型；
    机器学习模型至少基于所述得到的像输出预测信息。
  2. 根据权利要求1所述的方法，还包括：
    使用检测装置对量子比特进行检测，得到检测数据；
    将检测数据输入所述机器学习模型；
    所述机器学习模型基于测试数据和所述得到的像输出预测信息。
  3. 根据权利要求2所述的方法，还包括：
    获取量子比特的设计信息；
    将量子比特的设计信息输入所述机器学习模型；
    所述机器学习模型基于所述设计信息、所述测试数据和所述得到的像输出预测信息。
  4. 根据权利要求1-3中任一所述的方法，其中，所述的成像装置包括：扫描电子显微镜(SEM)，扫描隧道显微镜(STM)，红外成像装置，紫外成像装置，可见光范围内的成像装置。
  5. 根据权利要求4所述的方法，其中，使用成像装置在量子比特的制备过程中进行成像，所述制备过程包括多个步骤。
  6. 根据权利要求4所述的方法，其中，使用成像装置在量子比特制备完成后进行成像。
  7. 根据权利要求4所述的方法，其中，所述预测信息包括定量信息和定性信息。
  8. 根据权利要求7所述的方法，其中，所述定量信息包括以下至少一种：工作频率、相干时间、耦合强度。
  9. 根据权利要求7所述的方法，其中，所述的定性信息包括：量子比特的分级。
  10. 根据权利要求1-3中任一所述的方法，其中，所述的机器学习模型包括：卷积神经网络(CNN)、深度神经网络(DNN)、回归神经网络(RNN)。
  11. 根据权利要求1所述的方法，还包括：
    使用检测装置对量子比特进行检测；
    将测试数据输入所述机器学习模型；
    所述机器学习模型基于测试数据和所述得到的像输出预测信息。
  12. 一种量子比特检测系统，包括：
    测试模块，所述测试模块包括成像装置，所述成像装置配置为对量子比特进行成像；
    预测模块，所述预测模块和所述测试模块通信连接，所述预测模块包括机器学习模型，所述机器学习模型配置为至少基于得到的像输出预测信息。
  13. 根据权利要求12所述的量子比特检测系统，其中，所述测试模块还包括检测装置，所述检测装置配置为对量子比特进行检测以得到检测数据。
  14. 根据权利要求12所述的量子比特检测系统，其中，所述的成像装置包括：扫描电子显微镜(SEM)，扫描隧道显微镜(STM)，红外成像装置，紫外成像装置，可见光范围内的成像装置。
  15. 根据权利要求12所述的量子比特检测系统，其中，所述的机器学习模型包括：卷积神经网络(CNN)、深度神经网络(DNN)、回归神经网络(RNN)。
  16. 根据权利要求12所述的量子比特检测系统，还包括：交互模块和监控模块，所述监控模块和所述预测模块通信连接，所述交互模块和所述预测模块以及所述监控模块通信连接。
  17. 根据权利要求12所述的量子比特检测系统，还包括：测试控制模块，所述测试控制模块位于所述测试模块和所述预测模块之间，所述测试控制模块和所述测试模块以及所述预测模块通信连接，所述测试控制模块包括：成像数据产生模块、成像数据读取模块、量子比特测试产生模块、量子比特测试数据读取模块。
  18. 根据权利要求12所述的量子比特检测系统，其中，所述预测模块包括：量子比特缺陷判定和性质预测模块。
  19. 一种量子比特检测装置，包括：处理器和非暂态存储介质，所述非暂态存储介质存储有指令集，所述指令集被处理器执行时实现：
    用于使成像装置对量子比特进行成像的装置；
    用于将得到的像输入机器学习模型的装置；
    用于使机器学习模型至少基于所述得到的像输出预测信息的装置。
PCT/CN2019/103246 2018-09-05 2019-08-29 量子比特检测系统及检测方法 WO2020048375A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
JP2021510187A JP2021535483A (ja) 2018-09-05 2019-08-29 量子ビット検出システム及び検出方法
EP19858129.0A EP3848860A4 (en) 2018-09-05 2019-08-29 QUANTUM BIT DETECTION AND DETECTION SYSTEM
US17/189,719 US20210182727A1 (en) 2018-09-05 2021-03-02 Qubit detection system and detection method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811029930.4 2018-09-05
CN201811029930.4A CN110879105B (zh) 2018-09-05 2018-09-05 量子比特检测系统及检测方法

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/189,719 Continuation US20210182727A1 (en) 2018-09-05 2021-03-02 Qubit detection system and detection method

Publications (1)

Publication Number Publication Date
WO2020048375A1 true WO2020048375A1 (zh) 2020-03-12

Family

ID=69722173

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/103246 WO2020048375A1 (zh) 2018-09-05 2019-08-29 量子比特检测系统及检测方法

Country Status (6)

Country Link
US (1) US20210182727A1 (zh)
EP (1) EP3848860A4 (zh)
JP (1) JP2021535483A (zh)
CN (2) CN114264372A (zh)
TW (1) TWI803659B (zh)
WO (1) WO2020048375A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115840126A (zh) * 2021-09-18 2023-03-24 合肥本源量子计算科技有限责任公司 一种量子比特频率获取方法、量子测控系统和量子计算机

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11334667B1 (en) 2020-01-17 2022-05-17 Wells Fargo Bank, N.A. Systems and methods for disparate quantum computing threat detection
US11366897B1 (en) * 2020-01-17 2022-06-21 Wells Fargo Bank, N.A. Systems and methods for layered quantum computing detection
US11915325B2 (en) * 2021-06-09 2024-02-27 Bank Of America Corporation Quantum enabled resource activity investigation and response tool
US11829229B2 (en) 2022-04-26 2023-11-28 Hamilton Sundstrand Corporation Apparatus and method for diagnosing no fault failure found in electronic systems

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170330101A1 (en) * 2016-05-13 2017-11-16 Microsoft Technology Licensing, Llc Training a quantum optimizer
CN107612688A (zh) * 2017-09-30 2018-01-19 上海交通大学 基于机器学习的cvqkd实时性能优化方法及系统
WO2018106556A1 (en) * 2016-12-08 2018-06-14 Microsoft Technology Licensing, Llc Tomography and generative data modeling via quantum boltzmann training
CN108291878A (zh) * 2015-11-17 2018-07-17 科磊股份有限公司 单一图像检测
CN108369668A (zh) * 2015-10-16 2018-08-03 D-波系统公司 用于创建和使用量子玻尔兹曼机的系统和方法

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6078359A (en) * 1996-07-15 2000-06-20 The Regents Of The University Of California Vacuum compatible miniature CCD camera head
WO2002027653A2 (en) * 2000-09-26 2002-04-04 Robert Raussendorf Method for quantum computing
US20060056630A1 (en) * 2004-09-13 2006-03-16 Zimmer Vincent J Method to support secure network booting using quantum cryptography and quantum key distribution
JP5105410B2 (ja) * 2007-05-31 2012-12-26 独立行政法人科学技術振興機構 核磁気共鳴撮像システム及び撮像方法
US20160104073A1 (en) * 2012-12-05 2016-04-14 The United States Of America As Represented By The Secretary Of Commerce Radiation Suppression of Superconducting Quantum Bits Using a Conductive Plane
US9530873B1 (en) * 2013-01-28 2016-12-27 Sandia Corporation Semiconductor adiabatic qubits
WO2018069785A1 (en) * 2016-10-12 2018-04-19 Semiconductor Energy Laboratory Co., Ltd. Semiconductor device and system using the same
CN107300629B (zh) * 2017-07-31 2019-07-09 清华大学 扫描探针标定方法
US11262785B2 (en) * 2018-01-05 2022-03-01 University Of Maryland, College Park Methods and apparatuses for identifying and controlling quantum emitters in a quantum system
US12001918B2 (en) * 2018-01-18 2024-06-04 Google Llc Classification using quantum neural networks
CN108257132A (zh) * 2018-03-05 2018-07-06 南方医科大学 一种基于机器学习的ct图像质量评估的方法
US11536879B2 (en) * 2018-07-30 2022-12-27 IonQ, Inc. Elliptical beam design using cylindrical optics

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108369668A (zh) * 2015-10-16 2018-08-03 D-波系统公司 用于创建和使用量子玻尔兹曼机的系统和方法
CN108291878A (zh) * 2015-11-17 2018-07-17 科磊股份有限公司 单一图像检测
US20170330101A1 (en) * 2016-05-13 2017-11-16 Microsoft Technology Licensing, Llc Training a quantum optimizer
WO2018106556A1 (en) * 2016-12-08 2018-06-14 Microsoft Technology Licensing, Llc Tomography and generative data modeling via quantum boltzmann training
CN107612688A (zh) * 2017-09-30 2018-01-19 上海交通大学 基于机器学习的cvqkd实时性能优化方法及系统

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115840126A (zh) * 2021-09-18 2023-03-24 合肥本源量子计算科技有限责任公司 一种量子比特频率获取方法、量子测控系统和量子计算机
CN115840126B (zh) * 2021-09-18 2024-04-05 本源量子计算科技（合肥）股份有限公司 一种量子比特频率获取方法、量子测控系统和量子计算机

Also Published As

Publication number Publication date
TWI803659B (zh) 2023-06-01
TW202036392A (zh) 2020-10-01
CN110879105A (zh) 2020-03-13
CN114264372A (zh) 2022-04-01
EP3848860A4 (en) 2021-11-17
JP2021535483A (ja) 2021-12-16
US20210182727A1 (en) 2021-06-17
CN110879105B (zh) 2022-03-01
EP3848860A1 (en) 2021-07-14

Similar Documents

Publication Publication Date Title
WO2020048375A1 (zh) 量子比特检测系统及检测方法
US20210272010A1 (en) Systems and methods for interacting with a quantum computing system
US11907675B2 (en) Generating training datasets for training neural networks
CN108764281A (zh) 一种基于半监督自步学习跨任务深度网络的图像分类方法
JP2021504792A (ja) 量子分類器としての浅い回路のためのシステム、コンピュータ実装方法およびコンピュータ・プログラム
US20190095400A1 (en) Analytic system to incrementally update a support vector data description for outlier identification
EP4111373A1 (en) Robust artificial intelligence inference in edge computing devices
Mostaeen et al. Clonecognition: machine learning based code clone validation tool
Du et al. DCGAN based data generation for process monitoring
Cheng et al. Tell model where to attend: Improving interpretability of aspect-based sentiment classification via small explanation annotations
US20240135160A1 (en) System and method for efficient analyzing and comparing slice-based machine learn models
Rajendran et al. Data mining when each data point is a network
WO2015070759A1 (zh) 用于测试的大规模数据生成方法和装置
CN116775425A (zh) 一种基于并行自动编码器的少样本日志异常检测方法
CN116502714A (zh) 一种端到端的知识图谱生成方法
US20220207381A1 (en) Computer-readable recording medium having stored therein vector estimating program, apparatus for estimating vector, and method for estimating vector
US10366078B2 (en) Data reduction methods, systems, and devices
Fan et al. High-frequency keywords to predict defects for android applications
Pinilla et al. Structure-Aware Minor-Embedding for Machine Learning in Quantum Annealing Processors
Yang et al. The approach of graphical user interface testing guided by bayesian model
Wu et al. An Outlier Recognition Method Based on Improved CUSUM for GPS Time Series
Bentivegna Identifying Extreme Regimes in Climate-Scale Digital Twins: a Roadmap
Li et al. Learning Quantum Distributions Based on Normalizing Flow
CN118428375A (zh) 一种突发事件自动感知的方法及设备
KR20240078684A (ko) 이종 데이터 융합 방법 및 장치

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19858129

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021510187

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2019858129

Country of ref document: EP

Effective date: 20210406