Search

Optical imaging systems are inherently limited in their resolution due to the point spread function (PSF), which applies a static, yet spatially-varying, convolution to the image. This degradation can be addressed via Convolutional Neural Networks (CNNs), particularly through deblurring techniques. However, current solutions face certain limitations in efficiently computing spatially-varying convolutions. In this paper we propose CoordGate, a novel lightweight module that uses a multiplicative gate and a coordinate encoding network to enable efficient computation of spatially-varying convolutions in CNNs. CoordGate allows for selective amplification or attenuation of filters based on their spatial position, effectively acting like a locally connected neural network. The effectiveness of the CoordGate solution is demonstrated within the context of U-Nets and applied to the challenging problem of image deblurring. The experimental results show that CoordGate outperforms conventional approaches, offering a more robust and spatially aware solution for CNNs in various computer vision applications.

The next generation of high-power lasers enables repetition of experiments at orders of magnitude higher frequency than was possible using the prior generation. Facilities requiring human intervention between laser repetitions need to adapt in order to keep pace with the new laser technology. A distributed networked control system can enable laboratory-wide automation and feedback control loops. These higher-repetition-rate experiments will create enormous quantities of data. A consistent approach to managing data can increase data accessibility, reduce repetitive data-software development, and mitigate poorly organized metadata. An opportunity arises to share knowledge of improvements to control and data infrastructure currently being undertaken. We compare platforms and approaches to state-of-the-art control systems and data management at high-power laser facilities, and we illustrate these topics with case studies from our community., Comment: In press at High Power Laser Science and Engineering (2023), in special issue on "Future Control Systems and Machine Learning at High Power Laser Facilities." This Arxiv version is formatted for easy readability (single-column, 1.3x line spacing). Document navigation is available from the PDF sidebar, under "Bookmarks". 39 pages including references, 8 figures

All-optical Compton sources combine laser wakefield accelerators and intense scattering pulses to generate ultrashort bursts of backscattered radiation. The scattering pulse plays the role of a short-period undulator in which relativistic electrons oscillate and emit x-ray radiation. To date, most of the working laser-plasma accelerators operate preferably at energies of a few hundreds of MeV and the Compton sources developed so far produce radiation in the range from hundreds of keV to a few MeV. However, for such applications as medical imaging and tomography the relevant energy range is 10-100 keV. In this article, we discuss different scattering geometries for the generation of X-rays in this range. Through numerical simulations, we study the influence of electron beam parameters on the backscattered photons. We find that the spectral bandwidth remains constant for beams of the same emittance regardless of the scattering geometry. A shallow interaction angle of 30 degrees or less seems particularly promising for imaging applications given parameters of existing laser-plasma accelerators. Finally, we discuss the influence of the radiation properties for potential applications in medical imaging and non-destructive testing.

Presented is a novel way to combine snapshot compressive imaging and lateral shearing interferometry in order to capture the spatio-spectral phase of an ultrashort laser pulse in a single shot. A deep unrolling algorithm is utilised for the snapshot compressive imaging reconstruction due to its parameter efficiency and superior speed relative to other methods, potentially allowing for online reconstruction. The algorithm's regularisation term is represented using neural network with 3D convolutional layers, to exploit the spatio-spectral correlations that exist in laser wavefronts. Compressed sensing is not typically applied to modulated signals, but we demonstrate its success here. Furthermore, we train a neural network to predict the wavefronts from a lateral shearing interferogram in terms of Zernike polynomials, which again increases the speed of our technique without sacrificing fidelity. This method is supported with simulation-based results. While applied to the example of lateral shearing interferometry, the methods presented here are generally applicable to a wide range of signals, including Shack-Hartmann-type sensors. The results may be of interest beyond the context of laser wavefront characterization, including within quantitative phase imaging.

Laser-plasma physics has developed rapidly over the past few decades as high-power lasers have become both increasingly powerful and more widely available. Early experimental and numerical research in this field was restricted to single-shot experiments with limited parameter exploration. However, recent technological improvements make it possible to gather an increasing amount of data, both in experiments and simulations. This has sparked interest in using advanced techniques from mathematics, statistics and computer science to deal with, and benefit from, big data. At the same time, sophisticated modeling techniques also provide new ways for researchers to effectively deal with situations in which still only sparse amounts of data are available. This paper aims to present an overview of relevant machine learning methods with focus on applicability to laser-plasma physics, including its important sub-fields of laser-plasma acceleration and inertial confinement fusion.

Beam parameter optimization in accelerators involves multiple, sometimes competing objectives. Condensing these individual objectives into a single figure of merit unavoidably results in a bias towards particular outcomes, in absence of prior knowledge often in a non-desired way. Finding an optimal objective definition then requires operators to iterate over many possible objective weights and definitions, a process that can take many times longer than the optimization itself. A more versatile approach is multi-objective optimization, which establishes the trade-off curve or Pareto front between objectives. Here we present the first results on multi-objective Bayesian optimization of a simulated laser-plasma accelerator. We find that multi-objective optimization reaches comparable performance to its single-objective counterparts while allowing for instant evaluation of entirely new objectives. This dramatically reduces the time required to find appropriate objective definitions for new problems. Additionally, our multi-objective, multi-fidelity method reduces the time required for an optimization run by an order of magnitude. It does so by dynamically choosing simulation resolution and box size, requiring fewer slow and expensive simulations as it learns about the Pareto-optimal solutions from fast low-resolution runs. The techniques demonstrated in this paper can easily be translated into many different computational and experimental use cases beyond accelerator optimization.

The recent advent of deep artificial neural networks has resulted in a dramatic increase in performance for object classification and detection. While pre-trained with everyday objects, we find that a state-of-the-art object detection architecture can very efficiently be fine-tuned to work on a variety of object detection tasks in a high-power laser laboratory. In this manuscript, three exemplary applications are presented. We show that the plasma waves in a laser-plasma accelerator can be detected and located on the optical shadowgrams. The plasma wavelength and plasma density are estimated accordingly. Furthermore, we present the detection of all the peaks in an electron energy spectrum of the accelerated electron beam, and the beam charge of each peak is estimated accordingly. Lastly, we demonstrate the detection of optical damage in a high-power laser system. The reliability of the object detector is demonstrated over one thousand laser shots in each application. Our study shows that deep object detection networks are suitable to assist online and offline experiment analysis, even with small training sets. We believe that the presented methodology is adaptable yet robust, and we encourage further applications in high-power laser facilities regarding the control and diagnostic tools, especially for those involving image data.

In the pursuit of efficient optimization of expensive-to-evaluate systems, this paper investigates a novel approach to Bayesian multi-objective and multi-fidelity (MOMF) optimization. Traditional optimization methods, while effective, often encounter prohibitively high costs in multi-dimensional optimizations of one or more objectives. Multi-fidelity approaches offer potential remedies by utilizing multiple, less costly information sources, such as low-resolution simulations. However, integrating these two strategies presents a significant challenge. We suggest the innovative use of a trust metric to support simultaneous optimization of multiple objectives and data sources. Our method modifies a multi-objective optimization policy to incorporate the trust gain per evaluation cost as one objective in a Pareto optimization problem, enabling simultaneous MOMF at lower costs. We present and compare two MOMF optimization methods: a holistic approach selecting both the input parameters and the trust parameter jointly, and a sequential approach for benchmarking. Through benchmarks on synthetic test functions, our approach is shown to yield significant cost reductions - up to an order of magnitude compared to pure multi-objective optimization. Furthermore, we find that joint optimization of the trust and objective domains outperforms addressing them in sequential manner. We validate our results using the use case of optimizing laser-plasma acceleration simulations, demonstrating our method's potential in Pareto optimization of high-cost black-box functions. Implementing these methods in existing Bayesian frameworks is simple, and they can be readily extended to batch optimization. With their capability to handle various continuous or discrete fidelity dimensions, our techniques offer broad applicability in solving simulation problems in fields such as plasma physics and fluid dynamics.

We report results on all-optical Thomson scattering intercepting the acceleration process in a laser wakefield accelerator. We show that the pulse collision position can be detected using transverse shadowgraphy which also facilitates alignment. As the electron beam energy is evolving inside the accelerator, the emitted spectrum changes with the scattering position. Such a configuration could be employed as accelerator diagnostic as well as reliable setup to generate x-rays with tunable energy.

Recent progress in laser-driven plasma acceleration now enables the acceleration of electrons to several gigaelectronvolts. Taking advantage of these novel accelerators, ultra-short, compact and spatially coherent X-ray sources called betatron radiation have been developed and applied to high-resolution imaging. However, the scope of the betatron sources is limited by a low energy efficiency and a photon energy in the 10's of kiloelectronvolt range, which for example prohibits the use of these sources for probing dense matter. Here, based on three-dimensional particle-in-cell simulations, we propose an original hybrid scheme that combines a low-density laser-driven plasma accelerator with a high-density beam-driven plasma radiator, and thereby considerably increases the photon energy and the radiated energy of the betatron source. The energy efficiency is also greatly improved, with about 1% of the laser energy transferred to the radiation, and the gamma-ray photon energy exceeds the megaelectronvolt range when using a 15 J laser pulse. This high-brilliance hybrid betatron source opens the way to a wide range of applications requiring MeV photons, such as the production of medical isotopes with photo-nuclear reactions, radiography of dense objects in the defense or industrial domains and imaging in nuclear physics., Comment: 5 pages, 4 figures

Betatron x-ray source from laser plasma interaction combines high brightness, few femtosecond duration and broad band energy spectrum. However, despite these unique features the Betatron source has a crippling drawback preventing its use for applications. Its properties significantly vary shot-to-shot and none of the developments performed so far resolved this problem. In this letter we present a simple method that allows to produce stable and bright Betatron x-ray beams. In addition, we demonstrate that this scheme provides polarized and easily tunable radiation. Experimental results show that the pointing stability is better than 10% of the beam divergence, with flux fluctuation of the order of 20% and a polarization degree reaching up to 80%

Laser-plasma technology promises a drastic reduction of the size of high energy electron accelerators. It could make free electron lasers available to a broad scientific community, and push further the limits of electron accelerators for high energy physics. Furthermore the unique femtosecond nature of the source makes it a promising tool for the study of ultra-fast phenomena. However, applications are hindered by the lack of suitable lens to transport this kind of high-current electron beams, mainly due to their divergence. Here we show that this issue can be solved by using a laser-plasma lens, in which the field gradients are five order of magnitude larger than in conventional optics. We demonstrate a reduction of the divergence by nearly a factor of three, which should allow for an efficient coupling of the beam with a conventional beam transport line.