Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2005.07949 · Physical Review Letters

Machine Learning-Based Classification of Vector Vortex Beams

Taira Giordani, Alessia Suprano, Emanuele Polino, Francesca Acanfora, Luca Innocenti, Alessandro Ferraro, Mauro Paternostro, Nicolò Spagnolo, Fabio Sciarrino

Published Apr 20, 2020• 146 citations
arXivHTMLPDF

Abstract

Structured light is attracting significant attention for its diverse applications in both classical and quantum optics. The so-called vector vortex beams display peculiar properties in both contexts due to the nontrivial correlations between optical polarization and orbital angular momentum. Here we demonstrate a new, flexible experimental approach to the classification of vortex vector beams. We first describe a platform for generating arbitrary complex vector vortex beams inspired to photonic quantum walks. We then exploit recent machine learning methods-namely, convolutional neural networks and principal component analysis-to recognize and classify specific polarization patterns. Our study demonstrates the significant advantages resulting from the use of machine learning-based protocols for the construction and characterization of high-dimensional resources for quantum protocols.

DOI
10.1103/physrevlett.124.160401

Related papers

Engineering and characterization of high-dimensional states

Alessia Suprano, Taira Giordani, Emanuele Polino +10 more

The capability to engineer and characterize high dimensional states has become a crucial request in the quantum information field. The quantum walk dynamics proved to be a suitable resource for developing general quantumstate engineering protocols. Here, we experimentally verified the flexibility of an engineering protocol based on a one-dimensional quantum walk in the Orbital Angular Momentum (OAM). Although this degree of freedom has found several applications in the quantum information field, extract the information stored in them appears to be difficult. Therefore, we employ machine learning protocols to classify and characterize particularly structured beams endowed with a not uniform distribution of the polarization on the transverse plane. Moreover, we prove that by modeling the engineering process through a refined model it is possible to improve the performances of measurement techniques such as holographic projection and machine-learning based classification. These results represent a further investigation in the manipulation and detection of OAM modes coupling the photonics platforms with machine-learning protocols.

Quantum Computing, Communication, and Simulation•Mar 4, 2021•1 citation

Entangled vector vortex beams

Vincenzo D’Ambrosio, Gonzalo Carvacho, Francesco Graffitti +4 more

Light beams having a vectorial field structure, or polarization, that varies over the transverse profile and a central optical singularity are called vector vortex (VV) beams and may exhibit specific properties such as focusing into ``light needles'' or rotation invariance. VV beams have already found applications in areas ranging from microscopy to metrology, optical trapping, nano-optics, and quantum communication. Individual photons in such beams exhibit a form of single-particle quantum entanglement between different degrees of freedom. On the other hand, the quantum states of two photons can be also entangled with each other. Here, we combine these two concepts and demonstrate the generation of quantum entanglement between two photons that are both in VV states: a form of entanglement between two complex vectorial fields. This result may lead to quantum-enhanced applications of VV beams as well as to quantum information protocols fully exploiting the vectorial features of light.

arXiv:1507.08887•Physical Review A•Sep 30, 2016•114 citations

Experimental Property Reconstruction in a Photonic Quantum Extreme Learning Machine

Alessia Suprano, Danilo Zia, Luca Innocenti +10 more

Recent developments have led to the possibility of embedding machine learning tools into experimental platforms to address key problems, including the characterization of the properties of quantum states. Leveraging on this, we implement a quantum extreme learning machine in a photonic platform to achieve resource-efficient and accurate characterization of the polarization state of a photon. The underlying reservoir dynamics through which such input state evolves is implemented using the coined quantum walk of high-dimensional photonic orbital angular momentum and performing projective measurements over a fixed basis. We demonstrate how the reconstruction of an unknown polarization state does not need a careful characterization of the measurement apparatus and is robust to experimental imperfections, thus representing a promising route for resource-economic state characterization.

arXiv:2308.04543•Physical Review Letters•Apr 16, 2024•54 citations

Quantum walks and topological phenomena with structured light

Alessio D’Errico

The manipulation of the spatial structure of a light beam has many application in both classical and quantum physics. The possibility to exploit high dimensional degrees of freedom carried by a light beam can be employed, among the various applications, for simulating the dynamics on quantum particles in multidimensional spaces. Coupling these degrees of freedom, like the orbital angular momentum or the transverse linear momentum, with the polarization of light (also associated with the spin) allows to implement quantum walks on one and two dimensional lattices. This work presents a series of experiments where these implementations of photonic quantum walks were realized exploiting patterned liquid crystal devices. In particular, the experimental setup allows to investigate interesting effects of the non-trivial topological features of the simulated processes, and test methods for the experimental measurement of topological invariants. This research has also led to the introduction of a novel method of characterizing unknown structured light beams and to the study of the stability of polarization singularities in laser beams.

•Dec 3, 2018•0 citations

Experimental realization of a quantum classification: Bell state measurement via machine learning

Qingyuan Wu, Zhe Meng, Xiao-Xiao Chen +3 more

The Bell state is a crucial resource for the realization of quantum information tasks, and when combined with orbital angular momentum (OAM), it enables a high-dimensional Hilbert space, which is essential for high-capacity quantum communication. In this study, we demonstrate the recognition of OAM Bell states using interference patterns generated by a classical light source and a single-photon source from a Sagnac interferometer-based OAM Bell state evolution device. The interference patterns exhibit a one-to-one correspondence with the input Bell states, providing conclusive evidence for the full recognition of OAM Bell states. Furthermore, we introduce machine learning to the field of Bell state recognition by proposing a neural network model capable of accurately recognizing higher order single-photon OAM Bell states, even in the undersampling case. In particular, the model’s training set includes interference patterns of OAM Bell states generated by classical light sources, yet it is able to recognize single-photon OAM Bell states with high accuracy, without relying on quantum resources during training. Our innovative application of neural networks to the recognition of single-photon OAM Bell states not only circumvents the resource consumption and experimental difficulties associated with quantum light sources but also facilitates the study of OAM-based quantum information.

APL Machine Learning•Sep 1, 2023•2 citations

Machine-learning-assisted state and gate engineering for quantum technologies

Luca Innocenti

In this thesis, we discuss applications of machine learning to quantum information science. The interface between these two fields has been the subject of much research, recently, driven by the many successes of machine learning for diverse pattern recognition tasks. The work reported in this Thesis addresses precisely the potential of such an interdisciplinary line of research and illustrates how machine learning provides a valuable add-on to standard techniques used in the context of quantum technologies. Specifically, we will first study the problem of devising time-independent dynamics implementing target quantum operations. This involves a challenging optimisation task, which is hard to tackle with standard numerical tools. We demonstrate how the use of supervised learning algorithms can dramatically speed-up this task. We then consider the tasks of engineering and characterising quantum states in large Hilbert spaces. Motivated by strong experimental reasons, we consider explicitly the embodiment of such multi-dimensional quantum systems provided by the orbital angular momentum and polarisation degrees of freedom of light. We devise a protocol involving quantum walks to implement arbitrary states by only making use of the possibility to couple polarisation and orbital angular momentum through relatively recent technological advances in the field of linear optics. This protocol relies solely on the properties of quantum walk dynamics, and is therefore applicable to different types of architectures. We then present an experimental demonstration of this engineering strategy, and consider the issue of assessing the quality of the states thus generated. We show how different machine learning algorithms, including supervised and unsupervised learning ones, are able to tackle this problem and provide useful information in realistic experimental conditions.

•Oct 1, 2026•0 citations
Structured light is attracting significant attention for its diverse applications in both classical and quantum optics. The so-called vector vortex beams display peculiar properties in both contexts due to the nontrivial correlations between optical polarization and orbital angular momentum. Here we demonstrate a new, flexible experimental approach to the classification of vortex vector beams. We first describe a platform for generating arbitrary complex vector vortex beams inspired to photonic quantum walks. We then exploit recent machine learning methods-namely, convolutional neural networks and principal component analysis-to recognize and classify specific polarization patterns. Our study demonstrates the significant advantages resulting from the use of machine learning-based protocols for the construction and characterization of high-dimensional resources for quantum protocols.
The capability to engineer and characterize high dimensional states has become a crucial request in the quantum information field. The quantum walk dynamics proved to be a suitable resource for developing general quantumstate engineering protocols. Here, we experimentally verified the flexibility of an engineering protocol based on a one-dimensional quantum walk in the Orbital Angular Momentum (OAM). Although this degree of freedom has found several applications in the quantum information field, extract the information stored in them appears to be difficult. Therefore, we employ machine learning protocols to classify and characterize particularly structured beams endowed with a not uniform distribution of the polarization on the transverse plane. Moreover, we prove that by modeling the engineering process through a refined model it is possible to improve the performances of measurement techniques such as holographic projection and machine-learning based classification. These results represent a further investigation in the manipulation and detection of OAM modes coupling the photonics platforms with machine-learning protocols.
Light beams having a vectorial field structure, or polarization, that varies over the transverse profile and a central optical singularity are called vector vortex (VV) beams and may exhibit specific properties such as focusing into ``light needles'' or rotation invariance. VV beams have already found applications in areas ranging from microscopy to metrology, optical trapping, nano-optics, and quantum communication. Individual photons in such beams exhibit a form of single-particle quantum entanglement between different degrees of freedom. On the other hand, the quantum states of two photons can be also entangled with each other. Here, we combine these two concepts and demonstrate the generation of quantum entanglement between two photons that are both in VV states: a form of entanglement between two complex vectorial fields. This result may lead to quantum-enhanced applications of VV beams as well as to quantum information protocols fully exploiting the vectorial features of light.
Recent developments have led to the possibility of embedding machine learning tools into experimental platforms to address key problems, including the characterization of the properties of quantum states. Leveraging on this, we implement a quantum extreme learning machine in a photonic platform to achieve resource-efficient and accurate characterization of the polarization state of a photon. The underlying reservoir dynamics through which such input state evolves is implemented using the coined quantum walk of high-dimensional photonic orbital angular momentum and performing projective measurements over a fixed basis. We demonstrate how the reconstruction of an unknown polarization state does not need a careful characterization of the measurement apparatus and is robust to experimental imperfections, thus representing a promising route for resource-economic state characterization.
The manipulation of the spatial structure of a light beam has many application in both classical and quantum physics. The possibility to exploit high dimensional degrees of freedom carried by a light beam can be employed, among the various applications, for simulating the dynamics on quantum particles in multidimensional spaces. Coupling these degrees of freedom, like the orbital angular momentum or the transverse linear momentum, with the polarization of light (also associated with the spin) allows to implement quantum walks on one and two dimensional lattices. This work presents a series of experiments where these implementations of photonic quantum walks were realized exploiting patterned liquid crystal devices. In particular, the experimental setup allows to investigate interesting effects of the non-trivial topological features of the simulated processes, and test methods for the experimental measurement of topological invariants. This research has also led to the introduction of a novel method of characterizing unknown structured light beams and to the study of the stability of polarization singularities in laser beams.
The Bell state is a crucial resource for the realization of quantum information tasks, and when combined with orbital angular momentum (OAM), it enables a high-dimensional Hilbert space, which is essential for high-capacity quantum communication. In this study, we demonstrate the recognition of OAM Bell states using interference patterns generated by a classical light source and a single-photon source from a Sagnac interferometer-based OAM Bell state evolution device. The interference patterns exhibit a one-to-one correspondence with the input Bell states, providing conclusive evidence for the full recognition of OAM Bell states. Furthermore, we introduce machine learning to the field of Bell state recognition by proposing a neural network model capable of accurately recognizing higher order single-photon OAM Bell states, even in the undersampling case. In particular, the model’s training set includes interference patterns of OAM Bell states generated by classical light sources, yet it is able to recognize single-photon OAM Bell states with high accuracy, without relying on quantum resources during training. Our innovative application of neural networks to the recognition of single-photon OAM Bell states not only circumvents the resource consumption and experimental difficulties associated with quantum light sources but also facilitates the study of OAM-based quantum information.
In this thesis, we discuss applications of machine learning to quantum information science. The interface between these two fields has been the subject of much research, recently, driven by the many successes of machine learning for diverse pattern recognition tasks. The work reported in this Thesis addresses precisely the potential of such an interdisciplinary line of research and illustrates how machine learning provides a valuable add-on to standard techniques used in the context of quantum technologies. Specifically, we will first study the problem of devising time-independent dynamics implementing target quantum operations. This involves a challenging optimisation task, which is hard to tackle with standard numerical tools. We demonstrate how the use of supervised learning algorithms can dramatically speed-up this task. We then consider the tasks of engineering and characterising quantum states in large Hilbert spaces. Motivated by strong experimental reasons, we consider explicitly the embodiment of such multi-dimensional quantum systems provided by the orbital angular momentum and polarisation degrees of freedom of light. We devise a protocol involving quantum walks to implement arbitrary states by only making use of the possibility to couple polarisation and orbital angular momentum through relatively recent technological advances in the field of linear optics. This protocol relies solely on the properties of quantum walk dynamics, and is therefore applicable to different types of architectures. We then present an experimental demonstration of this engineering strategy, and consider the issue of assessing the quality of the states thus generated. We show how different machine learning algorithms, including supervised and unsupervised learning ones, are able to tackle this problem and provide useful information in realistic experimental conditions.