Search

This work presents a data-driven approach to estimating the sound absorption coefficient of an infinite porous slab using a neural network and a two-microphone measurement on a finite porous sample. A 1D-convolutional network predicts the sound absorption coefficient from the complex-valued transfer function between the sound pressure measured at the two microphone positions. The network is trained and validated with numerical data generated by a boundary element model using the Delany-Bazley-Miki model, demonstrating accurate predictions for various numerical samples. The method is experimentally validated with baffled rectangular samples of a fibrous material, where sample size and source height are varied. The results show that the neural network offers the possibility to reliably predict the in-situ sound absorption of a porous material using the traditional two-microphone method as if the sample were infinite. The normal-incidence sound absorption coefficient obtained by the network compares well with that obtained theoretically and in an impedance tube. The proposed method has promising perspectives for estimating the sound absorption coefficient of acoustic materials after installation and in realistic operational conditions., Comment: 41 pages, 8 figures

The enigmatic phenomenon of dark energy (DE) is regarded as the elusive entity driving the accelerated expansion of our Universe. A plausible candidate for DE is the non-zero Einstein Cosmological Constant $\Lambda_{E}$ manifested as a constant energy density of the vacuum, yet it seemingly defies gravitational effects. In this work, we interpret the non-zero $\Lambda_{E}$ through the lens of scale-invariant cosmology. We revisit the conformal scale factor $\lambda$ and its defining equations within the Scale-Invariant Vacuum (SIV) paradigm. Furthermore, we address the profound problem of the missing mass across galactic and extragalactic scales by deriving an MOND-like relation, $g \sim \sqrt{a_0\,g_N}$, within the SIV context. Remarkably, the values obtained for $\Lambda_{E}$ and the MOND fundamental acceleration, $a_0$, align with observed magnitudes, specifically, $a_0 \approx 10^{-10} \, \mathrm{m} \, \mathrm{s}^{-2}$ and $\Lambda_{E} \approx 1.8 \times 10^{-52} \, \mathrm{m}^{-2}$. Moreover, we propose a novel early dark energy term, $\tilde{T}_{\mu\nu} \sim \kappa H$, within the SIV paradigm, which holds potential relevance for addressing the Hubble tension. Keywords: cosmology; theory; dark energy; dark matter; MOND; Weyl integrable geometry., Comment: 13 pages, no figures

Graph neural networks (GNNs) have shown promise in learning the ground-state electronic properties of materials, subverting ab initio density functional theory (DFT) calculations when the underlying lattices can be represented as small and/or repeatable unit cells (i.e., molecules and periodic crystals). Realistic systems are, however, non-ideal and generally characterized by higher structural complexity. As such, they require large (10+ Angstroms) unit cells and thousands of atoms to be accurately described. At these scales, DFT becomes computationally prohibitive, making GNNs especially attractive. In this work, we present a strictly local equivariant GNN capable of learning the electronic Hamiltonian (H) of realistically extended materials. It incorporates an augmented partitioning approach that enables training on arbitrarily large structures while preserving local atomic environments beyond boundaries. We demonstrate its capabilities by predicting the electronic Hamiltonian of various systems with up to 3,000 nodes (atoms), 500,000+ edges, ~28 million orbital interactions (nonzero entries of H), and $\leq$0.55% error in the eigenvalue spectra. Our work expands the applicability of current electronic property prediction methods to some of the most challenging cases encountered in computational materials science, namely systems with disorder, interfaces, and defects., Comment: *Equal Contribution

Reservoir Computing was shown in recent years to be useful as efficient to learn networks in the field of time series tasks. Their randomized initialization, a computational benefit, results in drawbacks in theoretical analysis of large random graphs, because of which deterministic variations are an still open field of research. Building upon Next-Gen Reservoir Computing and the Temporal Convolution Derived Reservoir Computing, we propose a deterministic alternative to the higher-dimensional mapping therein, TCRC-LM and TCRC-CM, utilizing the parametrized but deterministic Logistic mapping and Chebyshev maps. To further enhance the predictive capabilities in the task of time series forecasting, we propose the novel utilization of the Lobachevsky function as non-linear activation function. As a result, we observe a new, fully deterministic network being able to outperform TCRCs and classical Reservoir Computing in the form of the prominent Echo State Networks by up to $99.99\%$ for the non-chaotic time series and $87.13\%$ for the chaotic ones.

Modern machine learning techniques have shown tremendous potential, especially for object detection on camera images. For this reason, they are also used to enable safety-critical automated processes such as autonomous drone flights. We present a study on object detection for Detect and Avoid, a safety critical function for drones that detects air traffic during automated flights for safety reasons. An ill-posed problem is the generation of good and especially large data sets, since detection itself is the corner case. Most models suffer from limited ground truth in raw data, \eg recorded air traffic or frontal flight with a small aircraft. It often leads to poor and critical detection rates. We overcome this problem by using inpainting methods to bootstrap the dataset such that it explicitly contains the corner cases of the raw data. We provide an overview of inpainting methods and generative models and present an example pipeline given a small annotated dataset. We validate our method by generating a high-resolution dataset, which we make publicly available and present it to an independent object detector that was fully trained on real data.

In recent years, data-driven deep learning models have gained significant interest in the analysis of turbulent dynamical systems. Within the context of reduced-order models (ROMs), convolutional autoencoders (CAEs) pose a universally applicable alternative to conventional approaches. They can learn nonlinear transformations directly from data, without prior knowledge of the system. However, the features generated by such models lack interpretability. Thus, the resulting model is a black-box which effectively reduces the complexity of the system, but does not provide insights into the meaning of the latent features. To address this critical issue, we introduce a novel interpretable CAE approach for high-dimensional fluid flow data that maintains the reconstruction quality of conventional CAEs and allows for feature interpretation. Our method can be easily integrated into any existing CAE architecture with minor modifications of the training process. We compare our approach to Proper Orthogonal Decomposition (POD) and two existing methods for interpretable CAEs. We apply all methods to three different experimental turbulent Rayleigh-B\'enard convection datasets with varying complexity. Our results show that the proposed method is lightweight, easy to train, and achieves relative reconstruction performance improvements of up to 6.4% over POD for 64 modes. The relative improvement increases to up to 229.8% as the number of modes decreases. Additionally, our method delivers interpretable features similar to those of POD and is significantly less resource-intensive than existing CAE approaches, using less than 2% of the parameters. These approaches either trade interpretability for reconstruction performance or only provide interpretability to a limited extend.

The continuous scaling of metal-oxide-semiconductor field-effect transistors (MOSFETs) has led to device geometries where charged carriers are increasingly confined to ever smaller channel cross sections. This development is associated with reduced screening of long-range Coulomb interactions. To accurately predict the behavior of such ultra-scaled devices, electron-electron (e-e) interactions must be explicitly incorporated in their quantum transport simulation. In this paper, we present an \textit{ab initio} atomistic simulation framework based on density functional theory, the non-equilibrium Green's function formalism, and the self-consistent GW approximation to perform this task. The implemented method is first validated with a carbon nanotube test structure before being applied to calculate the transfer characteristics of a silicon nanowire MOSFET in a gate-all-around configuration. As a consequence of e-e scattering, the energy and spatial distribution of the carrier and current densities both significantly change, while the on-current of the transistor decreases owing to the Coulomb repulsion between the electrons. Furthermore, we demonstrate how the resulting bandgap modulation of the nanowire channel as a function of the gate-to-source voltage could potentially improve the device performance. To the best of our knowledge, this study is the first one reporting large-scale atomistic quantum transport simulations of nano-devices under non-equilibrium conditions and in the presence of e-e interactions within the GW approximation.

The ability to concentrate sound energy with a tunable focal point is essential for a wide range of acoustic applications, offering precise control over the location and intensity of sound pressure maxima. However, conventional acoustic metalenses are typically passive, with fixed focal positions, limiting their versatility. A significant obstacle in achieving tunable sound wave focusing lies in the complexity of precise and programmable adjustments, which often require intricate mechanical or electronic systems. In this study, we present a theoretical and experimental investigation of a reconfigurable acoustic metalens based on a bistable origami design. The metalens comprises eight flexible origami units, each capable of switching between two stable equilibrium states, enabling local modulation of sound waves through two distinct reflection phases. The metalens can be locked into specific symmetric or asymmetric configurations by manually tailoring the origami units to settle either of the two states. Each configuration generates a unique phase profile, focusing sound energy at a specific point. This concept allows the focal spot to be dynamically reconfigured both on and off-axis. Furthermore, the approach introduces a simple yet effective mechanism for tuning sound energy concentration, offering a solution for flexible acoustic manipulation.

Band-to-band tunnelling field-effect transistors (TFETs) have long been considered as promising candidates for future low-power logic applications. However, fabricated TFETs rarely reach sub-60 mV/dec sub-threshold swings (SS) at room temperature. Previous theoretical studies identified Auger processes as possible mechanisms for the observed degradation of SS. Through first-principles quantum transport simulations incorporating carrier-carrier interactions within the Non-equilibrium Green's Function formalism and self-consistent GW approximation, we confirm here that Auger processes are indeed at least partly responsible for the poor performance of TFETs. Using a carbon nanotube TFET as testbed, we show that carrier-carrier scattering alone significantly increases the OFF-state current of these devices, thus worsening their sub-threshold behavior., Comment: 6 pages, 5 figures

We present a methodology for in situ Transmission Kikuchi Diffraction (TKD) tensile testing that enables nanoscale characterization of the evolution of complex plasticity mechanisms. By integrating a modified in situ scanning electron microscope nanoindenter with a microscale push-to-pull device and a conventional Electron Backscatter Diffraction (EBSD) detector, we achieved TKD measurements at high spatial resolution during mechanical deformation. A dedicated focused ion beam procedure was developed for site-specific specimen fabrication, including lift-out, thinning, and shaping into a dog-bone geometry. The methodology was demonstrated on two case studies: (i) a metastable $\beta$-Ti single crystal, on which we quantified the initiation and evolution of nanoscale twinning and stress-induced martensitic transformation, and (ii) a $CuAl/Al_2O_3$ nanolaminate, which showed nanoscale plasticity and twinning/detwinning in a complex microstructure. Overall, this approach provides a robust alternative to in situ EBSD and transmission electron microscopy testing, facilitating detailed analysis of deformation mechanisms at the nanoscale., Comment: 15 pages, 5 figures, accepted for publication in Scripta Materialia

We first briefly review the adventure of scale invariance in physics, from Galileo Galilei, Weyl, Einstein, and Feynman to the revival by Dirac (1973) and Canuto et al. (1977). We then gather concrete observational evidence that scale-invariant effects are present and measurable in astronomical objects spanning a vast range of masses (0.5 M$_{\odot} <$ M $< 10^{14}$ M$_{\odot}$) and an equally impressive range of spatial scales (0.01 pc $<$ r $<$ 1 Gpc). Scale invariance accounts for the observed excess in velocity in galaxy clusters with respect to the visible mass, the relatively flat/small slope of rotation curves in local galaxies, the observed steep rotation curves of high-redshift galaxies, and the excess of velocity in wide binary stars with separations above 3000 kau found in Gaia DR3. Last but not least, we investigate the effect of scale invariance on gravitational lensing. We show that scale invariance does not affect the geodesics of light rays as they pass in the vicinity of a massive galaxy. However, scale-invariant effects do change the inferred mass-to-light ratio of lens galaxies as compared to GR. As a result, the discrepancies seen in GR between the total lensing mass of galaxies and their stellar mass from photometry may be accounted for. This holds true both for lenses at high redshift like JWST-ER1 and at low redshift like in the SLACS sample. Of note is that none of the above observational tests require dark matter or any adjustable parameter to tweak the theory at any given mass or spatial scale., Comment: 25 pages, 6 figures, accepted in "Symmetry"

An equimolar NiCoFeCrGa high entropy alloy having dual-phase homogeneous components was studied, where the constituent phases exhibit distinct mechanical properties. Micropillars with various diameters were created from two differently heat treated samples, then they were compressed at slow strain rates, that revealed the material's limited sensitivity to size. On the other hand, increased strain rate sensitivity at high deformation speeds was observed, that differs substantially depending on the phase composition of the specimen. Dislocations within the two phases were studied by high resolution transmission electron microscopy and high angular resolution electron backscatter diffraction. The performed chemical analysis confirmed that slow cooling during casting create Cr-rich precipitates, that have significant impact on the global strength of the material.

The challenge in reconfigurable manipulation of sound waves using metasurfaces lies in achieving precise control over acoustic behavior while developing efficient and practical tuning methods for structural configurations. However, most studies on reconfigurable acoustic metasurfaces rely on cumbersome and time-consuming control systems. These approaches often struggle with fabrication techniques, as conventional methods face limitations such as restricted material choices, challenges in achieving complex geometries, and difficulties in incorporating flexible components. This paper proposes a novel approach for developing a reconfigurable metasurface inspired by the Kresling origami pattern, designed for programmable manipulation of acoustic waves at an operating frequency of 2000 Hz. The origami unit cell is fabricated using multi-material 3D printing technology, allowing for the simultaneous printing of two materials with different mechanical properties, thus creating a bistable origami-based structure. Through optimization, two equilibrium states achieve a reflection phase difference of {\pi} through the application of small axial force, F, or torque, T. Various configurations of the metasurface, generated from different combinations of these two equilibria, enable distinct reflective behaviors with switchable and programmable functionalities. The principle of this work simplifies the shaping of acoustic waves through a straightforward mechanical mechanism, eliminating the need for complex control systems and time-consuming adjustments. This innovative approach paves a novel and effective perspective for developing on-demand switchable and tunable devices across diverse fields, including electromagnetics, mechanics, and elastics, leveraging multi-material printing technology.

The prediction of time series is a challenging task relevant in such diverse applications as analyzing financial data, forecasting flow dynamics or understanding biological processes. Especially chaotic time series that depend on a long history pose an exceptionally difficult problem. While machine learning has shown to be a promising approach for predicting such time series, it either demands long training time and much training data when using deep Recurrent Neural Networks. Alternative, when using a Reservoir Computing approach it comes with high uncertainty and typically a high number of random initializations and extensive hyper-parameter tuning. In this paper, we focus on the Reservoir Computing approach and propose a new mapping of input data into the reservoir's state space. Furthermore, we incorporate this method in two novel network architectures increasing parallelizability, depth and predictive capabilities of the neural network while reducing the dependence on randomness. For the evaluation, we approximate a set of time series from the Mackey-Glass equation, inhabiting non-chaotic as well as chaotic behavior as well as the SantaFe Laser dataset and compare our approaches in regard to their predictive capabilities to Echo State Networks, Autoencoder connected Echo State Networks and Gated Recurrent Units. For the chaotic time series, we observe an error reduction of up to $85.45\%$ compared to Echo State Networks and $90.72\%$ compared to Gated Recurrent Units. Furthermore, we also observe tremendous improvements for non-chaotic time series of up to $99.99\%$ in contrast to the existing approaches.

Optical quantum communication technologies are making the prospect of unconditionally secure and efficient information transfer a reality. The possibility of generating and reliably detecting quantum states of light, with the further need of increasing the private data-rate is where most research efforts are focusing. The physical concept of entanglement is a solution guaranteeing the highest degree of security in device-independent schemes, yet its implementation and preservation over long communication links is hard to achieve. Lithium niobate-on-insulator has emerged as a revolutionising platform for high-speed classical telecommunication and is equally suited for quantum information applications owing to the large second-order nonlinearities that can efficiently produce entangled photon pairs. In this work, we generate maximally entangled quantum states in the time-bin basis using lithium niobate-on-insulator photonics at the fibre optics telecommunication wavelength, and reconstruct the density matrix by quantum tomography on a single photonic integrated circuit. We use on-chip periodically-poled lithium niobate as source of entangled qubits with a brightness of 242 MHz/mW and perform quantum tomography with a fidelity of 91.9+-1.0 %. Our results, combined with the established large electro-optic bandwidth of lithium niobate, showcase the platform as perfect candidate to realise fibre-coupled, high-speed time-bin quantum communication modules that exploit entanglement to achieve information security.

Recent studies of wide binary stars based on Gaia DR3 suggest that the relative orbital velocities of objects with separations s > 3'000 astronomical units are statistically larger than the standard Newtonian predictions. Obviously there is no Dark Matter halo arround binary stars that could be invoked to explain these high velocities. However, we explore the properties of two-body systems in the framework of scale invariant vacuum theory, focusing on the case of objects with extreme separations. In this regime, the additional acceleration term present in the modified Newton equation with scale invariance becomes important, and may even dominate the dynamical evolution at very low gravities. Comparisons with Gaia DR3 observations of wide binaries are performed and suggest that binaries with separations s > 3'000 astronomical units have experienced such an evolution for a few Gyr, accounting well for the observed velocity excesses., Comment: 17 pages, 3 figures

Resonant states underlie a variety of metastructures that exhibit remarkable capabilities for effective control of acoustic waves at subwavelength scales. The development of metamaterials relies on the rigorous mode engineering providing the implementation of the desired properties. At the same time, the application of metamaterials is still limited as their building blocks are frequently characterized by complicated geometry and can't be tuned easily. In this work, we consider a simple system of coupled Helmholtz resonators and study their properties associated with the tuning of coupling strength and symmetry breaking. We numerically and experimentally demonstrate the excitation of quasi-bound state in the continuum in the resonators placed in a free space and in a rectangular cavity. It is also shown that tuning the intrinsic losses via introducing porous inserts can lead to spectral splitting or merging of quasi-\textit{bound states in the continuum} and occurrence of \textit{exceptional points}. The obtained results will open new opportunities for the development of simple and easy-tunable metastructures based on Helmholtz resonances.

Intensity modulators are fundamental components for integrated photonics. From near infrared to visible spectral ranges, they find applications in optical communication and quantum technologies. In particular, they are required for the control and manipulation of atomic systems such as atomic clocks and quantum computers. Typical integrated electro-optic modulators operating at these wavelengths show high bandwidth and low voltage operation, but their extinction ratios are moderate. Here we present an integrated thin-film lithium niobate electro-optic modulator operating in the C-band, which uses a subsequent periodically poled waveguide to convert the modulated signal from 1536 nm to 768 nm using second harmonic generation. We demonstrate that the upconverted signal retains the characteristics of the modulated input signal, reaching a measured high bandwidth of 40 GHz. Due to the nature of the nonlinear process, it exhibits, with respect to the fundamental signal, a doubled extinction ratio of 46 dB, which is the highest recorded for near-infrared light on this platform., Comment: 4 pages, 3 figures

Generating and interfering non-classical states of light is fundamental to optical quantum information science and technology. Quantum photonic integrated circuits provide one pathway towards scalability by combining nonlinear sources of non-classical light and programmable circuits in centimeter-scale devices. The key requirements for quantum applications include efficient generation of indistinguishable photon-pairs and high-visibility programmable quantum interference. Here, we demonstrate a lithium niobate-on-insulator (LNOI) integrated photonic circuit that generates a two-photon path-entangled state, and a programmable interferometer for quantum interference. We generate entangled photons with $\sim2.3\times10^8$ pairs/s/mW brightness and perform quantum interference experiments on the chip with $96.8\pm3.6\%$ visibility. LNOI is an emerging photonics technology that has revolutionized high-speed modulators and efficient frequency conversion. Our results provide a path towards large-scale integrated quantum photonics including efficient photon-pair generation and programmable circuits for applications such as boson sampling and quantum communications.

We study the path of light rays passing near a massive object, in the context of the scale invariant equation of the geodesics first obtained by Dirac (1973). Using the exterior Schwarzschild solution for the metric, we derive the complete equations of the geodesics in the scale invariant context. We find that scale invariance introduces two additional terms to the Einstein term producing the deflection angle and that can potentially act over cosmological distances. Numerical integration of the scale-invariant geodesics, for the specific case of the z_L=1.94 lens galaxy in the extreme system JWST-ER1 (van Dokkum et al. 2023; Mercier et al. 2024) shows that the two additional terms introduce only negligible effects, typically 1E-06 of the Einstein term. We conclude that the lensing deflection angle derived in Einstein's General Relativity is essentially independent of the scale invariant effects and that the photon's geodesics remain unchanged. We also explore the possible origin of the differences in the mass estimates from lensing and photometry in JWST-ER1 and in the SLACS galaxies, differences which appear larger at higher redshifts. Scale invariance appears to release or even suppress the need for dark matter., Comment: 19 pages, 2 figures

Galaxy velocities in clusters, rotation curves of galaxies, and "vertical" oscillations in the Milky Way currently show too high velocities with respect to the masses thought to be involved. While these velocity excesses are currently interpreted as the consequence of dark matter, it can also be naturally explained as a consequence of scale invariant theory, which rests on a very simple first principle: the addition of a new fundamental symmetry. In the present work, the case of scale invariance, present in General Relativity and Maxwell equations for the empty space without charge and current, is considered. Cosmological models predict a rapid decrease of these effects with increasing mean density up to the critical density, where they totally disappear. Starting from the scale invariant geodesic equation by Dirac (1973), for which a demonstration by an action principle is presented, a modified Newton equation is derived. The solutions of this equation are applied to clusters of galaxies, galactic rotation at different redshifts and "vertical" motions in the Milky Way. In this new framework, the convergence of theoretical predictions and observations, in different gravitational systems, epochs, mass range and spatial scales, opens interesting perspectives that deserve to be explored further., Comment: 22 pages, 2 figures

The physics of stellar rotation plays a crucial role in the evolution of stars, their final fate and the properties of compact remnants. Diverse approaches have been adopted to incorporate the effects of rotation in stellar evolution models. This study seeks to explore the consequences of these various prescriptions for rotation on essential outputs of massive star models. We compute a grid of 15 and 60 M$_{\odot}$ stellar evolution models with the Geneva Stellar Evolution Code (GENEC), accounting for both hydrodynamical and magnetic instabilities induced by rotation. In both the 15 and 60 M$_{\odot}$ models, the choice of the vertical and horizontal diffusion coefficients for the non magnetic models strongly impacts the evolution of the chemical structure, but has a weak impact on the angular momentum transport and the rotational velocity of the core. In the 15 M$_{\odot}$ models, the choice of diffusion coefficient impacts the convective core size during the core H-burning phase, whether the model begins core He-burning as a blue or red supergiant and the core mass at the end of He-burning. In the 60 M$_{\odot}$ models, the evolution is dominated by mass loss and is less affected by the choice of diffusion coefficient. In the magnetic models, magnetic instability dominates the angular momentum transport and such models are found to be less mixed when compared to their rotating non-magnetic counterparts. Stellar models with the same initial mass, chemical composition, and rotation may exhibit diverse characteristics depending on the physics applied. By conducting thorough comparisons with observational features, we can ascertain which method(s) produce the most accurate results in different cases., Comment: Accepted in A&A, 18 pages

We present a summary of the main results within the Scale Invariant Vacuum (SIV) paradigm based on the Weyl Integrable Geometry (WIG) as an extension to the standard Einstein General Relativity (EGR). After a brief review of the mathematical framework, where we also highlight the connection between the weak-field SIV equations and the notion of un-proper time parametrization within the reparametrization paradigm, we continue with the main results related to early Universe; that is, applications to inflation, Big Bang Nucleosynthesis, and the growth of the density fluctuations within the SIV. In the late time Universe the applications of the SIV paradigm are related to scale-invariant dynamics of galaxies, MOND, dark matter, and the dwarf spheroidals where one can find MOND to be a peculiar case of the SIV theory. Finally, within the recent time epoch, we highlight that some of the change in the length-of-the-day (LOD), about 0.92 cm/yr, can be accounted for by SIV effects in the Earth-Moon system., Comment: 27 pages, 7 figures, 2 tables

On the basis of a general action principle, we revisit the scale invariant field equation using the co-tensor relations by Dirac (1973). This action principle also leads to an expression for the scale factor $\lambda$, which corresponds to the one derived from the gauging condition, which assumes that a macroscopic empty space is scale-invariant, homogeneous, and isotropic. These results strengthen the basis of the scale-invariant vacuum (SIV) paradigm. From the field and geodesic equations, we derive, in current time units (years, seconds), the Newton-like equation, the equations of the two-body problem, and its secular variations. In a two-body system, orbits very slightly expand, while the orbital velocity keeps constant during expansion. Interestingly enough, Kepler's third law is a remarkable scale-invariant property., Comment: 18 pages, no figures

A sharp lower bound for the injectivity radius in noncompact nonnegatively curved Riemannian manifolds involving their soul goes back to \v{S}arafutdinov. We generalize this bound to the setting of Alexandrov spaces. Our main theorem reads as follows. If the injectivity radius of an Alexandrov space of nonnegative curvature does not coincide with the one of its souls, then it is at least $ \pi K^{-1/2} $, where $ K $ is an upper curvature bound. We introduce the soul of Alexandrov spaces in some detail and compare two notions of injectivity radii., Comment: Comments are welcome

Gradient inversion attacks are an ubiquitous threat in federated learning as they exploit gradient leakage to reconstruct supposedly private training data. Recent work has proposed to prevent gradient leakage without loss of model utility by incorporating a PRivacy EnhanCing mODulE (PRECODE) based on variational modeling. Without further analysis, it was shown that PRECODE successfully protects against gradient inversion attacks. In this paper, we make multiple contributions. First, we investigate the effect of PRECODE on gradient inversion attacks to reveal its underlying working principle. We show that variational modeling introduces stochasticity into the gradients of PRECODE and the subsequent layers in a neural network. The stochastic gradients of these layers prevent iterative gradient inversion attacks from converging. Second, we formulate an attack that disables the privacy preserving effect of PRECODE by purposefully omitting stochastic gradients during attack optimization. To preserve the privacy preserving effect of PRECODE, our analysis reveals that variational modeling must be placed early in the network. However, early placement of PRECODE is typically not feasible due to reduced model utility and the exploding number of additional model parameters. Therefore, as a third contribution, we propose a novel privacy module -- the Convolutional Variational Bottleneck (CVB) -- that can be placed early in a neural network without suffering from these drawbacks. We conduct an extensive empirical study on three seminal model architectures and six image classification datasets. We find that all architectures are susceptible to gradient leakage attacks, which can be prevented by our proposed CVB. Compared to PRECODE, we show that our novel privacy module requires fewer trainable parameters, and thus computational and communication costs, to effectively preserve privacy., Comment: 14 pages (12 figures 6 tables) + 6 pages supplementary materials (6 tables). Under review. This work has been submitted to the IEEE for possible publication. arXiv admin note: substantial text overlap with arXiv:2208.04767

We summarize the main results within the Scale Invariant Vacuum (SIV) paradigm as related to the Weyl Integrable Geometry (WIG) as an extension to the standard Einstein General Relativity (EGR). After a short sketch of the mathematical framework, the main results until 2023 [1] are highlighted in relation to: the inflation within the SIV [2], the growth of the density fluctuations [3], the application of the SIV to scale-invariant dynamics of galaxies, MOND, dark matter, and the dwarf spheroidals [4],and the most recent results on the BBNS light-elements' abundances within the SIV [5]. Keywords: cosmology: theory, dark matter, dark energy, inflation, BBNS; galaxies: formation, rotation; Weyl integrable geometry; Dirac co-calculus., Comment: 10 pages, 3 figures, 2 tables; Comments and constructive suggestions are highly welcome. Presentation slides are available at ResearchGate.net via the link www.researchgate.net/publication/362388677; Contribution to the Relativity Theory conference at the Bauman Moscow State Technical University, Moscow, 4 July, 2023 (PIRT-2023 https://pirt.bmstu.ru/en/home-en/);. arXiv admin note: substantial text overlap with arXiv:2202.08412

LiDAR is crucial for robust 3D scene perception in autonomous driving. LiDAR perception has the largest body of literature after camera perception. However, multi-task learning across tasks like detection, segmentation, and motion estimation using LiDAR remains relatively unexplored, especially on automotive-grade embedded platforms. We present a real-time multi-task convolutional neural network for LiDAR-based object detection, semantics, and motion segmentation. The unified architecture comprises a shared encoder and task-specific decoders, enabling joint representation learning. We propose a novel Semantic Weighting and Guidance (SWAG) module to transfer semantic features for improved object detection selectively. Our heterogeneous training scheme combines diverse datasets and exploits complementary cues between tasks. The work provides the first embedded implementation unifying these key perception tasks from LiDAR point clouds achieving 3ms latency on the embedded NVIDIA Xavier platform. We achieve state-of-the-art results for two tasks, semantic and motion segmentation, and close to state-of-the-art performance for 3D object detection. By maximizing hardware efficiency and leveraging multi-task synergies, our method delivers an accurate and efficient solution tailored for real-world automated driving deployment. Qualitative results can be seen at https://youtu.be/H-hWRzv2lIY., Comment: Accepted for publication at IEEE Transactions on Intelligent Transportation Systems

The Scale Invariant Vacuum (SIV) paradigm is applied to the Big-Bang Nucleosynthesis using the known analytic expressions for the expansion factor $a$ and the plasma temperature $T$ as functions of the SIV time $\tau$ since the Big-Bang when $a(\tau=0)=0$. The results are compared to the known standard BBNS model as calculated with the PRIMAT code. Potential SIV-guided deviations from the local statistical equilibrium are explored. Overall, we find that smaller than usual baryon and non-zero dark matter content, by a factor of three to five times reduction, result in compatible to the standard reproduction of the light elements abundances., Comment: 12 pages, 3 figures, 4 tables

We present a graph-based model for multiple scattering of light in integrated lithium niobate on insulator (LNOI) networks, which describes an open network of single-mode integrated waveguides with tunable scattering at the network nodes. We first validate the model at small scale with experimental LNOI resonator devices and show consistent agreement between simulated and measured spectral data. Then, the model is used to demonstrate a novel platform for on-chip multiple scattering in large-scale optical networks up to few hundred nodes, with tunable scattering behaviour and tailored disorder. Combining our simple graph-based model with material properties of LNOI, this platform creates new opportunities to control randomness in large optical networks.

We present a robust and computationally efficient approach for assigning partial charges of atoms in molecules. The method is based on a hierarchical tree constructed from attention values extracted from a graph neural network (GNN), which was trained to predict atomic partial charges from accurate quantum-mechanical (QM) calculations. The resulting dynamic attention-based substructure hierarchy (DASH) approach provides fast assignment of partial charges with the same accuracy as the GNN itself, is software-independent, and can easily be integrated in existing parametrization pipelines as shown for the Open force field (OpenFF). The implementation of the DASH workflow, the final DASH tree, and the training set are available as open source / open data from public repositories.

Static analysis tools come in many forms andconfigurations, allowing them to handle various tasks in a (secure) development process: code style linting, bug/vulnerability detection, verification, etc., and adapt to the specific requirements of a software project, thus reducing the number of false positives.The wide range of configuration options poses a hurdle in their use for software developers, as the tools cannot be deployed out-of-the-box. However, static analysis tools only develop their full benefit if they are integrated into the software development workflow and used on regular. Vulnerability management should be integrated via version history to identify hotspots, for example. We present an analysis platform that integrates several static analysis tools that enable Git-based repositories to continuously monitor warnings across their version history. The framework is easily extensible with other tools and programming languages. We provide a visualization component in the form of a dashboard to display security trends and hotspots. Our tool can also be used to create a database of security alerts at a scale well-suited for machine learning applications such as bug or vulnerability detection.

Twinning is an important deformation mode in plastically deformed hexagonal close-packed materials. The extremely high twin growth rates at the nanoscale make atomistic simulations an attractive method for investigating the role of individual twin/matrix interfaces such as twin boundaries and basal-prismatic interfaces in twin growth kinetics. Unfortunately, there is no single framework that allows researchers to differentiate such interfaces automatically, neither in experimental in-situ transmission electron microscopy analysis images nor in atomistic simulations. Moreover, the presence of alloying elements introduces substantial noise to local atomic environments, making it nearly impossible to identify which atoms belong to which interface. Here, with the help of advanced machine learning methods, we provide a proof-of-concept way of using the local stress field distribution as an indicator for the presence of interfaces and for determining their types. We apply such an analysis to the growth of twin embryos in Mg-10 at.% Al alloys under constant stress and constant strain conditions, corresponding to two extremes of high and low strain rates, respectively. We discover that the kinetics of such growth is driven by high-energy basal-prismatic interfaces, in line with our experimental observations for pure Mg.

The scale invariant theory is preserving the fundamental physical properties of General Relativity, while enlarging the group of invariances subtending gravitation theory (Dirac1973; Canuto et al.1977). The Scale Invariant Vacuum (SIV) theory assumes, as gauging condition, that:"The macroscopic empty space is scale invariant, homogeneous and isotropic". Some basic properties in Weyl's Integrable Geometry and cotensor calculus are examined in relation with scalar-tensor theories. Possible scale invariant effects are strongly reduced by matter density, both at the cosmological and local levels. The weak feld limit of SIV tends to MOND, when the scale factor is taken as constant, an approximation valid (<1%) over the last 400 Myr. A better understanding of the a0-parameter is obtained: it corresponds to the equilibrium point of the Newtonian and SIV dynamical acceleration. Parameter a0 is not a universal constant, it depends on the density and age of the Universe. As MOND is doing, SIV theory avoids the call to dark matter, moreover the cosmological models predict accelerated expansion., Comment: 9 pages, 2 figures