© 2026 Aqora Quantum S.A.S.

Terms Privacy Legal Notice

arXiv:2106.04058 · Physical Review Letters

Extract the Degradation Information in Squeezed States with Machine Learning

Hsien-Yi Hsieh, Yi-Ru Chen, Hsun-Chung Wu, Hua Li Chen, Jingyu Ning, Y-C. Huang, Chien-Ming Wu, Ray-Kuang Lee

Published Feb 18, 2022• 32 citations

Abstract

In order to leverage the full power of quantum noise squeezing with unavoidable decoherence, a complete understanding of the degradation in the purity of squeezed light is demanded. By implementing machine-learning architecture with a convolutional neural network, we illustrate a fast, robust, and precise quantum state tomography for continuous variables, through the experimentally measured data generated from the balanced homodyne detectors. Compared with the maximum likelihood estimation method, which suffers from time-consuming and overfitting problems, a well-trained machine fed with squeezed vacuum and squeezed thermal states can complete the task of reconstruction of the density matrix in less than one second. Moreover, the resulting fidelity remains as high as 0.99 even when the antisqueezing level is higher than 20 dB. Compared with the phase noise and loss mechanisms coupled from the environment and surrounding vacuum, experimentally, the degradation information is unveiled with machine learning for low and high noisy scenarios, i.e., with the antisqueezing levels at 12 dB and 18 dB, respectively. Our neural network enhanced quantum state tomography provides the metrics to give physical descriptions of every feature observed in the quantum state with a single scan measurement just by varying the local oscillator phase from 0 to 2π and paves a way of exploring large-scale quantum systems in real time.

DOI: 10.1103/physrevlett.128.073604

Related papers

Machine learning for quantum state tomography: robust covariance matrix estimation for squeezed vacuum states with thermal noise

Juan Camilo Rodríguez, Hsien-Yi Hsieh, Hua Li Chen +3 more

We present a supervised machine learning-based method using convolutional neural networks to estimate the covariance matrix of Gaussian quantum states in the presence of thermal noise. Unlike computationally intensive density matrix reconstructions, our machine learning-based method allows for the reconstruction of impure squeezed vacuum states using sparse measurements of quadrature sequences based on a model employing a two-component state mixed together from thermal and squeezed thermal states. The method achieves high fidelity and precision, notably also at high squeezing levels, while offering an effective characterization of physical quantities and accurately estimating the covariance matrix. We benchmark our machine against experimental data of single-mode squeezed vacuum states, demonstrating its accuracy and capability to quantify experimental degradation to squeezing and purity. We experimentally verify that our covariance matrix estimation exhibits robustness to state degradation induced by thermal state admixtures. We provide a method for lightweight, compact, and complete representation of lab-generated Gaussian states and lay the foundation for extending real-time quantum state tomography for thermal multi-component Gaussian states to multi-mode systems.

Optics Express•Oct 31, 2025•0 citations

Direct Parameter Estimations from Machine Learning-Enhanced Quantum State Tomography

Hsien-Yi Hsieh, Jingyu Ning, Yi-Ru Chen +4 more

With the power to find the best fit to arbitrarily complicated symmetry, machine-learning (ML)-enhanced quantum state tomography (QST) has demonstrated its advantages in extracting complete information about the quantum states. Instead of using the reconstruction model in training a truncated density matrix, we develop a high-performance, lightweight, and easy-to-install supervised characteristic model by generating the target parameters directly. Such a characteristic model-based ML-QST can avoid the problem of dealing with a large Hilbert space, but cab keep feature extractions with high precision, capturing the underlying symmetry in data. With the experimentally measured data generated from the balanced homodyne detectors, we compare the degradation information about quantum noise squeezed states predicted by the reconstruction and characteristic models; both are in agreement with the empirically fitting curves obtained from the covariance method. Such a ML-QST with direct parameter estimations illustrates a crucial diagnostic toolbox for applications with squeezed states, from quantum information process, quantum metrology, advanced gravitational wave detectors, to macroscopic quantum state generation.

arXiv:2203.16385•Symmetry•Apr 25, 2022•25 citations

Neural Network Enhanced Single-Photon Fock State Tomography

Hsien-Yi Hsieh, Yi-Ru Chen, Jingyu Ning +7 more

Even though heralded single-photon sources have been generated routinely through the spontaneous parametric down conversion, vacuum and multiple photon states are unavoidably involved. With machine-learning, we report the experimental implementation of single-photon quantum state tomography by directly estimating target parameters. Compared to the Hanbury Brown and Twiss (HBT) measurements only with clicked events recorded, our neural network enhanced quantum state tomography characterizes the photon number distribution for all possible photon number states from the balanced homodyne detectors. By using the histogram-based architecture, a direct parameter estimation on the negativity in Wigner's quasi-probability phase space is demonstrated. Such a fast, robust, and precise quantum state tomography provides us a crucial diagnostic toolbox for the applications with single-photon Fock states and other non-Gaussisan quantum states.

arXiv:2405.02812•May 5, 2024•0 citations

Neural-network-enhanced Fock-state tomography

Hsien-Yi Hsieh, Yi-Ru Chen, Mi-Mi Huang +8 more

Even though heralded single-photon sources have been generated routinely through spontaneous parametric down conversion, vacuum and multiple photon states are unavoidably involved. With machine learning, we report the experimental implementation of photon number state tomography by directly estimating target parameters. Compared to the Hanbury Brown and Twiss measurements only with clicked events recorded, our neural-network-enhanced quantum-state tomography characterizes the photon number distribution for all possible photon number states from the balanced homodyne detectors. By using the histogram-based architecture, a direct parameter estimation on the negativity in Wigner's quasiprobability phase space is demonstrated. Such a fast, robust, and precise quantum-state tomography provides us a crucial diagnostic toolbox for the applications with incoherent mixture of Fock states.

Physical Review A•Nov 7, 2024•4 citations

Experimental quantum homodyne tomography via machine learning

E. S. Tiunov, V. V. Tiunova, A. E. Ulanov +2 more

Complete characterization of states and processes that occur within quantum devices is crucial for understanding and testing their potential to outperform classical technologies for communications and computing. However, solving this task with current state-of-the-art techniques becomes unwieldy for large and complex quantum systems. Here we realize and experimentally demonstrate a method for complete characterization of a quantum harmonic oscillator based on an artificial neural network known as the restricted Boltzmann machine. We apply the method to optical homodyne tomography and show it to allow full estimation of quantum states based on a smaller amount of experimental data compared to state-of-the-art methods. We link this advantage to reduced overfitting. Although our experiment is in the optical domain, our method provides a way of exploring quantum resources in a broad class of large-scale physical systems, such as superconducting circuits, atomic and molecular ensembles, and optomechanical systems.

arXiv:1907.06589•Optica•Apr 13, 2020•68 citations

Machine-learning enhanced quantum state tomography and quantum noise reduction to the advanced gravitational wave detectors

Ray‐Kuang Lee

With this talk, I will first illustrate the implementation of our machine-learning (ML) enhanced quantum state tomography (QST) for continuous variables, through the experimentally measured data generated from squeezed vacuum states, as an example of quantum machine learning. At the same time, as a collaborator for LIGO-VirgoKAGRA (LVK) gravitational wave network and Einstein Telescope, our plan to inject this squeezed vacuum field into the advanced gravitational wave detectors (GWD) will be introduced. Finally, I will report our recent progress in applying such a ML-QST as a crucial diagnostic toolbox for applications with squeezed states, from Wigner currents, optical cat state generation, and Bayesian estimation for GWD.

Optical and Quantum Sensing and Precision Metrology II•Mar 13, 2024•1 citation

In order to leverage the full power of quantum noise squeezing with unavoidable decoherence, a complete understanding of the degradation in the purity of squeezed light is demanded. By implementing machine-learning architecture with a convolutional neural network, we illustrate a fast, robust, and precise quantum state tomography for continuous variables, through the experimentally measured data generated from the balanced homodyne detectors. Compared with the maximum likelihood estimation method, which suffers from time-consuming and overfitting problems, a well-trained machine fed with squeezed vacuum and squeezed thermal states can complete the task of reconstruction of the density matrix in less than one second. Moreover, the resulting fidelity remains as high as 0.99 even when the antisqueezing level is higher than 20 dB. Compared with the phase noise and loss mechanisms coupled from the environment and surrounding vacuum, experimentally, the degradation information is unveiled with machine learning for low and high noisy scenarios, i.e., with the antisqueezing levels at 12 dB and 18 dB, respectively. Our neural network enhanced quantum state tomography provides the metrics to give physical descriptions of every feature observed in the quantum state with a single scan measurement just by varying the local oscillator phase from 0 to 2π and paves a way of exploring large-scale quantum systems in real time.

We present a supervised machine learning-based method using convolutional neural networks to estimate the covariance matrix of Gaussian quantum states in the presence of thermal noise. Unlike computationally intensive density matrix reconstructions, our machine learning-based method allows for the reconstruction of impure squeezed vacuum states using sparse measurements of quadrature sequences based on a model employing a two-component state mixed together from thermal and squeezed thermal states. The method achieves high fidelity and precision, notably also at high squeezing levels, while offering an effective characterization of physical quantities and accurately estimating the covariance matrix. We benchmark our machine against experimental data of single-mode squeezed vacuum states, demonstrating its accuracy and capability to quantify experimental degradation to squeezing and purity. We experimentally verify that our covariance matrix estimation exhibits robustness to state degradation induced by thermal state admixtures. We provide a method for lightweight, compact, and complete representation of lab-generated Gaussian states and lay the foundation for extending real-time quantum state tomography for thermal multi-component Gaussian states to multi-mode systems.

With the power to find the best fit to arbitrarily complicated symmetry, machine-learning (ML)-enhanced quantum state tomography (QST) has demonstrated its advantages in extracting complete information about the quantum states. Instead of using the reconstruction model in training a truncated density matrix, we develop a high-performance, lightweight, and easy-to-install supervised characteristic model by generating the target parameters directly. Such a characteristic model-based ML-QST can avoid the problem of dealing with a large Hilbert space, but cab keep feature extractions with high precision, capturing the underlying symmetry in data. With the experimentally measured data generated from the balanced homodyne detectors, we compare the degradation information about quantum noise squeezed states predicted by the reconstruction and characteristic models; both are in agreement with the empirically fitting curves obtained from the covariance method. Such a ML-QST with direct parameter estimations illustrates a crucial diagnostic toolbox for applications with squeezed states, from quantum information process, quantum metrology, advanced gravitational wave detectors, to macroscopic quantum state generation.

Even though heralded single-photon sources have been generated routinely through the spontaneous parametric down conversion, vacuum and multiple photon states are unavoidably involved. With machine-learning, we report the experimental implementation of single-photon quantum state tomography by directly estimating target parameters. Compared to the Hanbury Brown and Twiss (HBT) measurements only with clicked events recorded, our neural network enhanced quantum state tomography characterizes the photon number distribution for all possible photon number states from the balanced homodyne detectors. By using the histogram-based architecture, a direct parameter estimation on the negativity in Wigner's quasi-probability phase space is demonstrated. Such a fast, robust, and precise quantum state tomography provides us a crucial diagnostic toolbox for the applications with single-photon Fock states and other non-Gaussisan quantum states.

Even though heralded single-photon sources have been generated routinely through spontaneous parametric down conversion, vacuum and multiple photon states are unavoidably involved. With machine learning, we report the experimental implementation of photon number state tomography by directly estimating target parameters. Compared to the Hanbury Brown and Twiss measurements only with clicked events recorded, our neural-network-enhanced quantum-state tomography characterizes the photon number distribution for all possible photon number states from the balanced homodyne detectors. By using the histogram-based architecture, a direct parameter estimation on the negativity in Wigner's quasiprobability phase space is demonstrated. Such a fast, robust, and precise quantum-state tomography provides us a crucial diagnostic toolbox for the applications with incoherent mixture of Fock states.

Complete characterization of states and processes that occur within quantum devices is crucial for understanding and testing their potential to outperform classical technologies for communications and computing. However, solving this task with current state-of-the-art techniques becomes unwieldy for large and complex quantum systems. Here we realize and experimentally demonstrate a method for complete characterization of a quantum harmonic oscillator based on an artificial neural network known as the restricted Boltzmann machine. We apply the method to optical homodyne tomography and show it to allow full estimation of quantum states based on a smaller amount of experimental data compared to state-of-the-art methods. We link this advantage to reduced overfitting. Although our experiment is in the optical domain, our method provides a way of exploring quantum resources in a broad class of large-scale physical systems, such as superconducting circuits, atomic and molecular ensembles, and optomechanical systems.

With this talk, I will first illustrate the implementation of our machine-learning (ML) enhanced quantum state tomography (QST) for continuous variables, through the experimentally measured data generated from squeezed vacuum states, as an example of quantum machine learning. At the same time, as a collaborator for LIGO-VirgoKAGRA (LVK) gravitational wave network and Einstein Telescope, our plan to inject this squeezed vacuum field into the advanced gravitational wave detectors (GWD) will be introduced. Finally, I will report our recent progress in applying such a ML-QST as a crucial diagnostic toolbox for applications with squeezed states, from Wigner currents, optical cat state generation, and Bayesian estimation for GWD.