Category: Machine Learning

Fully Programmable Spatial Photonic Ising Machine by Focal Plane Division

https://arxiv.org/abs/2410.10689

Ising machines are an emerging class of hardware that promises ultrafast and energy-efficient solutions to NP-hard combinatorial optimization problems. Spatial photonic Ising machines (SPIMs) exploit optical computing in free space to accelerate the computation, showcasing parallelism, scalability, and low power consumption. However, current SPIMs can implement only a restricted class of problems. This partial programmability is a critical limitation that hampers their benchmark. Achieving full programmability of the device while preserving its scalability is an open challenge. Here, we report a fully programmable SPIM achieved through a novel operation method based on the division of the focal plane. In our scheme, a general Ising problem is decomposed into a set of Mattis Hamiltonians, whose energies are simultaneously computed optically by measuring the intensity on different regions of the camera sensor. Exploiting this concept, we experimentally demonstrate the computation with high success probability of ground-state solutions of up to 32-spin Ising models on unweighted maximum cut graphs with and without ferromagnetic bias. Simulations of the hardware prove a favorable scaling of the accuracy with the number of spins. Our fully programmable SPIM enables the implementation of many quadratic unconstrained binary optimization problems, further establishing SPIMs as a leading paradigm in non von Neumann hardware.

https://mathstodon.xyz/@nonlinearxwaves/113310235718756468

Deep Learning Enabled Transmission of Full-Stokes Polarization Images Through Complex Media

Polarization images offer crucial functionalities across multiple scientific domains, providing access to physical information beyond conventional measures such as intensity, phase, and spectrum of light. However, the challenge of transmitting polarization images through complex media has restricted their application in optical communication and imaging. Here, a novel approach utilizing deep learning for the transmission of full-Stokes polarization images through scattering media is presented. It is demonstrated that any input polarization image can be reconstructed in a single shot by employing only an intensity sensor. By supervised training of a deep neural network, high-accuracy full-Stokes reconstruction is achieved from the speckle pattern detected by an intensity camera. Leveraging the deep learning based polarization decoder, a polarization-colored encoding scheme is devised to enable increased-capacity data transmission through disordered channels. Fast, wavelength-independent, on-chip, polarization imaging in complex media enables the utilization of polarization-structured light in multimode fibres and opaque materials, unlocking new possibilities in optical communication, cryptography, and quantum technology.

https://doi.org/10.1002/lpor.202400626

Tensorial flow of mosaic beams in PRL !

https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.132.243801

Optical beams with nonuniform polarization offer enhanced capabilities for information transmission, boasting increased capacity, security, and resilience. These beams possess vectorial features that are spatially organized within localized three-dimensional regions, forming tensors that can be harnessed across a spectrum of applications spanning quantum physics, imaging, and machine learning. However, when subjected to the effect of the transmission channel, the tensorial propagation leads to a loss of data integrity due to the entanglement of spatial and polarization degrees of freedom. The challenge of quantifying this spatial-polarization coupling poses a significant obstacle to the utilization of vector beams in turbulent environments, multimode fibers, and disordered media. Here, we introduce and experimentally investigate mosaic vector beams, which consist of localized polarization tesserae that propagate in parallel, demonstrating accurate measurement of their behavior as they traverse strongly disordered channels and decoding their polarization structure in single-shot experiments. The resultant transmission tensor empowers polarization-based optical communication and imaging in complex media. These findings also hold promise for photonic machine learning, where the engineering of tensorial flow can enable optical computing with high throughput.

Efficient Computation Using Spatial-Photonic Ising Machines: Utilizing Low-Rank and Circulant Matrix Constraints

https://arxiv.org/abs/2406.01400

We explore the potential of spatial-photonic Ising machines (SPIMs) to address computationally intensive Ising problems that employ low-rank and circulant coupling matrices. Our results indicate that the performance of SPIMs is critically affected by the rank and precision of the coupling matrices. By developing and assessing advanced decomposition techniques, we expand the range of problems SPIMs can solve, overcoming the limitations of traditional Mattis-type matrices. Our approach accommodates a diverse array of coupling matrices, including those with inherently low ranks, applicable to complex NP-complete problems. We explore the practical benefits of low-rank approximation in optimization tasks, particularly in financial optimization, to demonstrate the real-world applications of SPIMs. Finally, we evaluate the computational limitations imposed by SPIM hardware precision and suggest strategies to optimize the performance of these systems within these constraints.