Search

In this work, we study the problem where a group of mobile agents needs to reach a set of goal locations, but it does not matter which agent reaches a specific goal. Unlike most of the existing works on this topic that typically assume the existence of the centralized planner (or controller) and limit the agents' moves to a predefined graph of locations and transitions between them, in this work we focus on the decentralized scenarios, when each agent acts individually relying only on local observations/communications and is free to move in arbitrary direction at any time. Our iterative approach involves agents individually selecting goals, exchanging them, planning paths, and at each time step choose actions that balance between progressing along the paths and avoiding collisions. The proposed method is shown to be complete under specific assumptions on how agents progress towards their current goals, and our empirical evaluation demonstrates its superiority over a baseline decentralized navigation approach in success rate (i.e. is able to solve more problem instances under a given time limit) and a comparison with the centralized TSWAP algorithm reveals its efficiency in minimizing trajectory lengths for mission accomplishment., Comment: This is a pre-print of the paper accepted to ICR 2024. It contains 16 pages, 5 figures, 1 table

Planning safe paths in 3D workspace for high DoF robotic systems, such as manipulators, is a challenging problem, especially when the environment is populated with the dynamic obstacles that need to be avoided. In this case the time dimension should be taken into account that further increases the complexity of planning. To mitigate this issue we suggest to combine safe-interval path planning (a prominent technique in heuristic search) with the randomized planning, specifically, with the bidirectional rapidly-exploring random trees (RRT-Connect) - a fast and efficient algorithm for high-dimensional planning. Leveraging a dedicated technique of fast computation of the safe intervals we end up with an efficient planner dubbed SI-RRT. We compare it with the state of the art and show that SI-RRT consistently outperforms the competitors both in runtime and solution cost. Our implementation of SI-RRT is publicly available at https://github.com/PathPlanning/ManipulationPlanning-SI-RRT

Compressed streak imaging (CSI), introduced in 2014, has proven to be a powerful imaging technology for recording ultrafast phenomena such as light propagation and fluorescence lifetimes at over 150 trillion frames per second. Despite these achievements, CSI has faced challenges in detecting subtle intensity fluctuations in slow-moving, continuously illuminated objects. This limitation, largely attributable to high streak compression and motion blur, has curtailed broader adoption of CSI in applications such as cellular fluorescence microscopy. To address these issues and expand the utility of CSI, we present a novel encoding strategy, termed two-axis compressed streak imaging (TACSI) that results in significant improvements to the reconstructed image fidelity. TACSI introduces a second scanning axis which shuttles a conjugate image of the object with respect to the coded aperture. The moving image decreases the streak compression ratio and produces a flash and shutter phenomenon that reduces coded aperture motion blur, overcoming the limitations of current CSI technologies. We support this approach with an analytical model describing the two-axis streak compression ratio, along with both simulated and empirical measurements. As proof of concept, we demonstrate the ability of TACSI to measure rapid variations in cell membrane potentials using voltage-sensitive dye, which were previously unattainable with conventional CSI. This method has broad implications for high-speed photography, including the visualization of action potentials, muscle contractions, and enzymatic reactions that occur on microsecond and faster timescales using fluorescence microscopy., Comment: 29 pages, 11 figures

We study a path planning problem where the possible move actions are represented as a finite set of motion primitives aligned with the grid representation of the environment. That is, each primitive corresponds to a short kinodynamically-feasible motion of an agent and is represented as a sequence of the swept cells of a grid. Typically heuristic search, i.e. A*, is conducted over the lattice induced by these primitives (lattice-based planning) to find a path. However due to the large branching factor such search may be inefficient in practice. To this end we suggest a novel technique rooted in the idea of searching over the grid cells (as in vanilla A*) simultaneously fitting the possible sequences of the motion primitives into these cells. The resultant algorithm, MeshA*, provably preserves the guarantees on completeness and optimality, on the one hand, and is shown to notably outperform conventional lattice-based planning (x1.5 decrease in the runtime), on the other hand. Moreover, we suggest an additional pruning technique that additionally decreases the search space of MeshA*. The resultant planner is combined with the regular A* to retain completeness and is shown to further increase the search performance at the cost of negligible decrease of the solution quality.

The increasing demand for processing large volumes of data for machine learning models has pushed data bandwidth requirements beyond the capability of traditional von Neumann architecture. In-memory computing (IMC) has recently emerged as a promising solution to address this gap by enabling distributed data storage and processing at the micro-architectural level, significantly reducing both latency and energy. In this paper, we present the IMPACT: InMemory ComPuting Architecture Based on Y-FlAsh Technology for Coalesced Tsetlin Machine Inference, underpinned on a cutting-edge memory device, Y-Flash, fabricated on a 180 nm CMOS process. Y-Flash devices have recently been demonstrated for digital and analog memory applications, offering high yield, non-volatility, and low power consumption. The IMPACT leverages the Y-Flash array to implement the inference of a novel machine learning algorithm: coalesced Tsetlin machine (CoTM) based on propositional logic. CoTM utilizes Tsetlin automata (TA) to create Boolean feature selections stochastically across parallel clauses. The IMPACT is organized into two computational crossbars for storing the TA and weights. Through validation on the MNIST dataset, IMPACT achieved 96.3% accuracy. The IMPACT demonstrated improvements in energy efficiency, e.g., 2.23X over CNN-based ReRAM, 2.46X over Neuromorphic using NOR-Flash, and 2.06X over DNN-based PCM, suited for modern ML inference applications., Comment: 27 Pages, 14 Figures, 6 Tables

The spin properties of charge carriers confined in CuCl semiconductor nanocrystals (NCs) of different sizes (radius from 1.8 nm up to 28 nm) crystallized in a glass matrix are studied experimentally and theoretically. By means of photoluminescence, spin-flip Raman scattering, time-resolved Faraday ellipticity, and time-resolved differential transmission performed at temperatures in the range of $1.6 - 120$ K at magnetic fields up to 8 T, comprehensive information on the Land\'e $g$-factors as well as the population and spin dynamics is received. The spin signals are contributed by confined electrons with a $g$-factor close to 2, which shows a weak increase with decreasing NC size, i.e. increasing electron confinement energy. We revisit the theory of exciton confinement as a whole in spherical NCs within the six-band valence band model in order to describe the size dependence of the $Z_3$ and $Z_{1,2}$ exciton energies in CuCl NCs. We demonstrate theoretically that the stronger increase of the $Z_{1,2}$ energy transitions with decreasing radius can be explained by the strong absorption from the excited exciton state caused by strong heavy hole-light hole mixing in the exciton. The parameters of the six-band Hamiltonian describing both the exciton and hole kinetic energies are estimated from the comparison of the calculated and experimental size dependences of the exciton transitions. A theoretical model of the size-dependent Land\'e $g$-factors for electron and hole confined in spherical NCs of semiconductors with negative spin-orbit splitting of the valence band is developed., Comment: 33 pages Main file and 9 pages SI

Topological boundary modes utilizing valley mode waveguides have opened opportunities in, for instance, the design of high transmission waveguides with tolerance to geometrical defects and sharp bends. Applications of these waveguides include linear computational processes and the emulation of logic gates using linear structures, among other scenarios. Here we present the design of a 6-port junction that exhibits equal power splitting to three other ports when excited at single port with no reflections. In studying this structure, a scattering matrix is extracted at telecom wavelengths (around 1550 nm). The linearity of the system along with the scattering matrix are exploited to produce linear operations such as routing of information considering two incident signals or multiple signals applied from different ports. Our work may be exploited to analytically design larger networks without the need of computationally expensive trial and error numerical methods., Comment: 27 pages, 5 Figures, 83 References

Ab initio calculation of low energy scattering of electrons (muons) off hydrogen (muonic hydrogen) are performed on the basis of Faddeev-Merkuriev (FM) equations. The explicit contribution of induced dipole interaction in the asymptotic behavior of the wave function components has been incorporated into FM formalism. Elastic and inelastic cross sections have been calculated with high energy resolution in the vicinity of $n=2,3$ exited states thresholds of respective atoms. The Gailitis-Dumburg oscillations are discovered in some of calculated cross sections., Comment: 11 pages, 4 figures

This paper proposes a novel approach to semantic ontology alignment using contextual descriptors. A formalization was developed that enables the integration of essential and contextual descriptors to create a comprehensive knowledge model. The hierarchical structure of the semantic approach and the mathematical apparatus for analyzing potential conflicts between concepts, particularly in the example of "Transparency" and "Privacy" in the context of artificial intelligence, are demonstrated. Experimental studies showed a significant improvement in ontology alignment metrics after the implementation of contextual descriptors, especially in the areas of privacy, responsibility, and freedom & autonomy. The application of contextual descriptors achieved an average overall improvement of approximately 4.36%. The results indicate the effectiveness of the proposed approach for more accurately reflecting the complexity of knowledge and its contextual dependence., Comment: Ontology alignment, contextual descriptors, semantic matching, knowledge representation, essential descriptors, ontology integration, hierarchical structure, semantic heterogeneity, ethical AI

Magnetars are neutron stars with superstrong magnetic fields. Some of them (soft-gamma repeaters, SGRs) demonstrate gigantic flares which nature is still unclear. At decay phase of such flares one often observes quasi-periodic oscillations (QPOs) which are treated as stellar oscillations triggered by the flares. We study, for the first time, magneto-elastic oscillations of magnetars possessing toroidal magnetic fields confined in the stellar crust, without imposing axial symmetry of perturbations. We show that the Zeeman effect makes the oscillation spectrum much richer than for axially symmetric oscillations. The main properties of theoretical QPO spectra are discussed as well as their potential to interpret observations and explore magnetar physics., Comment: 15 pages, 9 figures, submitted to Phys. Rev. D

As part of the PIP-II project at Fermilab, a pre-production cryomodule featuring 325 MHz Single Spoke Resonator type 2 (SSR2) superconducting RF cavities is under construction. These SSR2 cavities are fabricated by industry partners and undergo initial cold testing at our collaborating institution, IJCLab in France, utilizing low-power coupler. Subsequently, the cavities are subjected to final qualification at Fermilab, complete with tuner and high-power coupler assemblies. This paper provides an overview of the ongoing efforts dedicated to high-power testing of jacketed SSR2 cavities in the Spoke Test Cryostat (STC) at Fermilab. Performance parameters obtained from these tests are presented, offering valuable insights into the cavities operational characteristics and readiness for integration into the PIP-II cryomodule., Comment: 32nd Linear Accelerator Conference (LINAC 2024)

Brillouin Light Scattering (BLS) spectroscopy is a non-invasive, non-contact, label-free optical technique that can provide information on the mechanical properties of a material on the sub-micron scale. Over the last decade it has seen increased applications in the life sciences, driven by the observed significance of mechanical properties in biological processes, the realization of more sensitive BLS spectrometers and its extension to an imaging modality. As with other spectroscopic techniques, BLS measurements not only detect signals characteristic of the investigated sample, but also of the experimental apparatus, and can be significantly affected by measurement conditions. The aim of this consensus statement is to improve the comparability of BLS studies by providing reporting recommendations for the measured parameters and detailing common artifacts. Given that most BLS studies of biological matter are still at proof-of-concept stages and use different--often self-built--spectrometers, a consensus statement is particularly timely to assure unified advancement., Comment: Main Text & Supplementary Text: 56 pages, 3 Figures, 2 Supplementary Figures, 1 Supplementary Table

Optimization of a trade-off between the number of speakers and their temporal variability (or session diversity) is crucial for the development of a speaker recognition system together with making the data collection process feasible from a time perspective. In this article, we provide the analysis of dependency between inter and intra speaker variability in training data for the modern neural network-based speaker recognition system using the VoxTube dataset for text-independent speaker recognition task. Besides, an auxiliary contribution of this work is a release of upload date metadata per utterance in a VoxTube dataset. We want this article to contribute to guidelines and best practices for collecting and filtering data from media hosting platforms to facilitate the efforts of researchers in developing speaker recognition systems.

The great variety of lead halide perovskite semiconductors represents an outstanding platform for studying crystal symmetry effects on the spin-dependent properties. Access to them is granted through optical orientation of exciton and carrier spins by circularly polarized photons. Here, the exciton spin polarization is investigated at $1.6$\,K cryogenic temperature in four lead halide perovskite crystals with different symmetries: (almost) cubic in FA$_{0.9}$Cs$_{0.1}$PbI$_{2.8}$Br$_{0.2}$ and FAPbBr$_3$, and orthorhombic in MAPbI$_3$ and CsPbBr$_3$. Giant optical orientation of 85\% is found for the excitons in FA$_{0.9}$Cs$_{0.1}$PbI$_{2.8}$Br$_{0.2}$, MAPbI$_3$, and CsPbBr$_3$, while it amounts to 20\% in FAPbBr$_3$. For all studied crystals, the optical orientation is robust to detuning of the laser photon energy from the exciton resonance, remaining constant for high energy detunings up to 0.3~eV, above which it continuously decreases to zero for detunings exceeding 1~eV. No acceleration of the spin relaxation for excitons with large kinetic energy is found in the cubic and orthorhombic crystals. This evidences the absence of the Dyakonov-Perel spin relaxation mechanism, which is based on the Rashba-Dresselhaus splitting of spin states at finite $k$-vectors. This indicates that the spatial inversion symmetry is maintained in perovskite crystals, independent of the cubic or orthorhombic phase.

We present an all-optical concept for measuring the electric field of light spanning from infrared to extreme ultraviolet with multi-petahertz detection bandwidth. Our approach employs a heterodyne detection of light produced by a highly nonlinear light-matter interaction gate. We establish a numerical model of a complex spectral response for unambiguous electric field extraction and benchmark it against the experiment. We show that the concept can be applied for measuring wavelengths down to about 60 nm with a high sensitivity and dynamic range of about 40 decibels. This opens up unprecedented perspectives for spatio-temporal electric field-resolved experiments and control of broadband dynamics of matter.

The recently proposed ToolkenGPT tool learning paradigm demonstrates promising performance but suffers from two major issues: first, it cannot benefit from tool documentation, and second, it often makes mistakes in whether to use a tool at all. We introduce Toolken+ that mitigates the first problem by reranking top $k$ tools selected by ToolkenGPT and the second problem with a special "Reject" option such that the model will generate a vocabulary token if "Reject" is ranked first. We demonstrate the effectiveness of Toolken+ on multistep numerical reasoning and tool selection tasks., Comment: EMNLP 2024 Findings

Autonomous navigation of a mobile robot is a challenging task which requires ability of mapping, localization, path planning and path following. Conventional mapping methods build a dense metric map like an occupancy grid, which is affected by odometry error accumulation and consumes a lot of memory and computations in large environments. Another approach to mapping is the usage of topological properties, e.g. adjacency of locations in the environment. Topological maps are less prone to odometry error accumulation and high resources consumption, and also enable fast path planning because of the graph sparsity. Based on this idea, we proposed NavTopo - a full navigation pipeline based on topological map and two-level path planning. The pipeline localizes in the graph by matching neural network descriptors and 2D projections of the input point clouds, which significantly reduces memory consumption compared to metric and topological point cloud-based approaches. We test our approach in a large indoor photo-relaistic simulated environment and compare it to a metric map-based approach based on popular metric mapping method RTAB-MAP. The experimental results show that our topological approach significantly outperforms the metric one in terms of performance, keeping proper navigational efficiency., Comment: This paper is published in proceedings of the 9th International Conference "Interactive Collaborative Robotics" (ICR 2024)

The Land\'e $g$-factor of charge carriers is a key parameter in spin physics controlling spin polarization and spin dynamics. In turn, it delivers information of the electronic band structure in vicinity of the band gap and its modification in nanocrystals provided by strong carrier confinement. The coherent spin dynamics of electrons and holes are investigated in CsPbI$_3$ perovskite nanocrystals with sizes varied from 4 to 16 nm by means of time-resolved Faraday ellipticity at the temperature of 6 K. The Land\'e $g$-factors of the charge carriers are evaluated through the Larmor spin precession in magnetic fields up to 430 mT across the spectral range from 1.69 to 2.25 eV, provided by variation of the nanocrystal size. The spectral dependence of the electron $g$-factor follows the model predictions when accounting for the mixing of the electronic bands with increasing confinement resulting from a decrease of the nanocrystal size. The spectral dependence of the hole $g$-factor, changing from $-0.19$ to $+1.69$, is considerably stronger than expected from the model. We analyze several mechanisms and conclude that none of them can be responsible for this difference. The renormalizations of the electron and hole $g$-factors roughly compensate each other, providing spectral independence for the bright exciton $g$-factor with a value of about $+2.2$., Comment: arXiv admin note: text overlap with arXiv:2407.21610

Holography is an essential technique of generating three-dimensional images. Recently, quantum holography with undetected photons (QHUP) has emerged as a groundbreaking method capable of capturing complex amplitude images. Despite its potential, the practical application of QHUP has been limited by susceptibility to phase disturbances, low interference visibility, and limited spatial resolution. Deep learning, recognized for its ability in processing complex data, holds significant promise in addressing these challenges. In this report, we present an ample advancement in QHUP achieved by harnessing the power of deep learning to extract images from single-shot holograms, resulting in vastly reduced noise and distortion, alongside a notable enhancement in spatial resolution. The proposed and demonstrated deep learning QHUP (DL-QHUP) methodology offers a transformative solution by delivering high-speed imaging, improved spatial resolution, and superior noise resilience, making it suitable for diverse applications across an array of research fields stretching from biomedical imaging to remote sensing. DL-QHUP signifies a crucial leap forward in the realm of holography, demonstrating its immense potential to revolutionize imaging capabilities and pave the way for advancements in various scientific disciplines. The integration of DL-QHUP promises to unlock new possibilities in imaging applications, transcending existing limitations and offering unparalleled performance in challenging environments.

Scattering obscures information carried by wave by producing a speckle pattern, posing a common challenge across various fields, including microscopy and astronomy. Traditional methods for extracting information from speckles often rely on significant physical assumptions, complex devices, or intricate algorithms. Recently, machine learning has emerged as a scalable and widely adopted tool for interpreting speckle patterns. However, most current machine learning techniques depend heavily on supervised training with extensive labeled datasets, which is problematic when labels are unavailable. To address this, we propose a strategy based on unsupervised learning for speckle recognition and evaluation, enabling to capture high-level information, such as object classes, directly from speckles without labeled data. By deriving invariant features from speckles, this method allows for the classification of speckles and facilitates diverse applications in image sensing. We experimentally validated our strategy through two significant applications: a noninvasive glucose monitoring system capable of differentiating time-lapse glucose concentrations, and a high-throughput communication system utilizing multimode fibers in dynamic environments. The versatility of this method holds promise for a broad range of far-reaching applications, including biomedical diagnostics, quantum network decoupling, and remote sensing.

Magnetars are neutron stars with superstrong magnetic fields which can exceed 1e15 G. Some magnetars (the so-called soft gamma-repeaters) demonstrate occasionally very powerful processes of energy release, which result in exceptionally strong flares of electromagnetic radiation. It is believed that these flares are associated with the presence of superstrong magnetic fields. Despite many hypotheses, the mechanism of these flares remains a mystery. In afterglows of the flares, one has often observed quasi-periodic oscillations (QPOs) of magnetar emission. They are interpreted as stellar vibrations, excited by the flares, which are useful for exploring the nature of magnetar activity. The incompleteness of theories employed to interpret magnetar QPOs is discussed., Comment: 9 pages, 5 figures. Published in Russian, English translation by the author; is expected to be published in English later

Compact conductively cooled SRF industrial linacs can provide unique parameters of the electron beam for industrial applications. (up to 10MeV, 1MW). For ERDC project we designed normal conducting RF injector with thermal RF gridded gun integrated in first cell of multi-cell cavities. For design of the RF gun we used MICHELLE software to simulate and optimize parameters of the beam. Output file was converted to ASTRA format and most beam dynamic simulations in multi-cell normal conducting cavity and cryomodule were performed by using ASTRA software. For cross-checking we compare results of MICHELLE and AS-TRA in first few cells. At the end of injector beam reach ~250keV energy which allow to trap bunch in acceleration regime without losses in TESLA like 1.3 GHz cavity. Short solenoid at the end of injector allow to regulate transverse beam size in cryomodule to match beam to extraction system and also reduce charge losses in accelerator., Comment: 32nd Linear Accelerator Conference (LINAC 2024)

The thermionic gridded gun is described which generates short electron bunches for further acceleration in a Nb3Sn conduction cooled SRF linac. The gun is built into the first cavity of the 250 keV injector [1] for the 20 kW, 10 MeV, 1.3 GHz CW conduction cooled one-cavity linac. The beam current is 2 mA. The RF gun design is presented as well as the results of perveance measurements, which are in a good agreement with the design parameters. The design of the RF resonator of the gun is presented also. The beam generated by the gun is matched to the injector to provide lack of current interception in the SRF cavity., Comment: 32nd Linear Accelerator Conference (LINAC 2024)

The coherent spin dynamics of holes are investigated for CsPbBr$_3$ and CsPb(Cl,Br)$_3$ perovskite nanocrystals in a glass matrix using the time-resolved Faraday rotation/ellipticity techniques. In an external magnetic field, pronounced Larmor spin precession of the hole spins is detected across a wide temperature range from 5 to 300 K. The hole Land\'e $g$-factor varies in the range of $0.8-1.5$, in which it increases with increasing high energy shift of the exciton due to enhanced confinement in small nanocrystals. The hole spin dephasing time decreases from 1 ns to 50 ps in this temperature range. Nuclear spin fluctuations have a pronounced impact on the hole spin dynamics. The hyperfine interaction of the holes with nuclear spins modifies their spin polarization decay and induces their spin precession in zero external magnetic field. The results can be well described by the model developed in Ref. 41, from which the hyperfine interaction energy of a hole spin with the nuclear spin fluctuation in range of $2-5$ $\mu$eV is evaluated.

Multi-agent pathfinding (MAPF) is a challenging computational problem that typically requires to find collision-free paths for multiple agents in a shared environment. Solving MAPF optimally is NP-hard, yet efficient solutions are critical for numerous applications, including automated warehouses and transportation systems. Recently, learning-based approaches to MAPF have gained attention, particularly those leveraging deep reinforcement learning. Following current trends in machine learning, we have created a foundation model for the MAPF problems called MAPF-GPT. Using imitation learning, we have trained a policy on a set of pre-collected sub-optimal expert trajectories that can generate actions in conditions of partial observability without additional heuristics, reward functions, or communication with other agents. The resulting MAPF-GPT model demonstrates zero-shot learning abilities when solving the MAPF problem instances that were not present in the training dataset. We show that MAPF-GPT notably outperforms the current best-performing learnable-MAPF solvers on a diverse range of problem instances and is efficient in terms of computation (in the inference mode).

In this paper we study a challenging variant of the multi-agent pathfinding problem (MAPF), when a set of agents must reach a set of goal locations, but it does not matter which agent reaches a specific goal - Anonymous MAPF (AMAPF). Current optimal and suboptimal AMAPF solvers rely on the existence of a centralized controller which is in charge of both target assignment and pathfinding. We extend the state of the art and present the first AMAPF solver capable of solving the problem at hand in a fully decentralized fashion, when each agent makes decisions individually and relies only on the local communication with the others. The core of our method is a priority and target swapping procedure tailored to produce consistent goal assignments (i.e. making sure that no two agents are heading towards the same goal). Coupled with an established rule-based path planning, we end up with a TP-SWAP, an efficient and flexible approach to solve decentralized AMAPF. On the theoretical side, we prove that TP-SWAP is complete (i.e. TP-SWAP guarantees that each target will be reached by some agent). Empirically, we evaluate TP-SWAP across a wide range of setups and compare it to both centralized and decentralized baselines. Indeed, TP-SWAP outperforms the fully-decentralized competitor and can even outperform the semi-decentralized one (i.e. the one relying on the initial consistent goal assignment) in terms of flowtime (a widespread cost objective in MAPF, Comment: This is a pre-print of the paper accepted to ECAI 2024. Its main body is similar the camera-ready version of the conference paper. In addition this pre-print contains Supplementary Material incorporating extended empirical results and analysis. It contains 10 pages, 8 figures, 4 tables

Reinforcement learning is a widely used approach to autonomous navigation, showing potential in various tasks and robotic setups. Still, it often struggles to reach distant goals when safety constraints are imposed (e.g., the wheeled robot is prohibited from moving close to the obstacles). One of the main reasons for poor performance in such setups, which is common in practice, is that the need to respect the safety constraints degrades the exploration capabilities of an RL agent. To this end, we introduce a novel learnable algorithm that is based on decomposing the initial problem into smaller sub-problems via intermediate goals, on the one hand, and respects the limit of the cumulative safety constraints, on the other hand -- SPEIS(Safe Policy Exploration Improvement via Subgoals). It comprises the two coupled policies trained end-to-end: subgoal and safe. The subgoal policy is trained to generate the subgoal based on the transitions from the buffer of the safe (main) policy that helps the safe policy to reach distant goals. Simultaneously, the safe policy maximizes its rewards while attempting not to violate the limit of the cumulative safety constraints, thus providing a certain level of safety. We evaluate SPEIS in a wide range of challenging (simulated) environments that involve different types of robots in two different environments: autonomous vehicles from the POLAMP environment and car, point, doggo, and sweep from the safety-gym environment. We demonstrate that our method consistently outperforms state-of-the-art competitors and can significantly reduce the collision rate while maintaining high success rates (higher by 80% compared to the best-performing methods)., Comment: 11 pages, 8 figures

Fermilab recently completed production and testing of 1.3 GHz cryomodules for the LCLS-II project. Each cryomodule consists of eight TESLA-shaped superconducting elliptical cavities equipped with two High Order Mode (HOM) coupler ports. Measurement of the HOM spectrum is part of the incoming quality control of cavities at room temperature and the final qualification cold test of cryomodules at the Cryomodule Test Facility (CMTF). In this paper we describe the procedure for measuring the HOM spectrum along with further data processing. Finally, we present accumulated statistics of individual HOM frequencies and quality factors related to various cavity vendors and discuss the possible contribution of HOMs to heat loads and beam dynamics., Comment: 32nd Linear Accelerator Conference (LINAC 2024)

The modern implementation of machine learning architectures faces significant challenges due to frequent data transfer between memory and processing units. In-memory computing, primarily through memristor-based analog computing, offers a promising solution to overcome this von Neumann bottleneck. In this technology, data processing and storage are located inside the memory. Here, we introduce a novel approach that utilizes floating-gate Y-Flash memristive devices manufactured with a standard 180 nm CMOS process. These devices offer attractive features, including analog tunability and moderate device-to-device variation; such characteristics are essential for reliable decision-making in ML applications. This paper uses a new machine learning algorithm, the Tsetlin Machine (TM), for in-memory processing architecture. The TM's learning element, Automaton, is mapped into a single Y-Flash cell, where the Automaton's range is transferred into the Y-Flash's conductance scope. Through comprehensive simulations, the proposed hardware implementation of the learning automata, particularly for Tsetlin machines, has demonstrated enhanced scalability and on-edge learning capabilities.

The coherent spin dynamics of electrons and holes in CsPbI3 perovskite nanocrystals in a glass matrix are studied by the time-resolved Faraday ellipticity technique in magnetic fields up to 430 mT across a temperature range from 6 K up to 120 K. The Land\'e g-factors and spin dephasing times are evaluated from the observed Larmor precession of electron and hole spins. The nanocrystal size in the three studied samples varies from about 8 to 16 nm, resulting in exciton transition varying from 1.69 to 1.78 eV at the temperature of 6 K, allowing us to study the corresponding energy dependence of the g-factors. The electron g-factor decreases with increasing confinement energy in the NCs as result of NC size reduction, and also with increasing temperature. The hole g-factor shows the opposite trend. A model analysis shows that the variation of g-factors with NC size arises from the transition energy dependence of the g-factors, which becomes strongly renormalized by temperature.

In this paper, we present Reshape Dimensions Network (ReDimNet), a novel neural network architecture for extracting utterance-level speaker representations. Our approach leverages dimensionality reshaping of 2D feature maps to 1D signal representation and vice versa, enabling the joint usage of 1D and 2D blocks. We propose an original network topology that preserves the volume of channel-timestep-frequency outputs of 1D and 2D blocks, facilitating efficient residual feature maps aggregation. Moreover, ReDimNet is efficiently scalable, and we introduce a range of model sizes, varying from 1 to 15 M parameters and from 0.5 to 20 GMACs. Our experimental results demonstrate that ReDimNet achieves state-of-the-art performance in speaker recognition while reducing computational complexity and the number of model parameters., Comment: Proceedings of Interspeech

Multi-agent reinforcement learning (MARL) has recently excelled in solving challenging cooperative and competitive multi-agent problems in various environments with, mostly, few agents and full observability. Moreover, a range of crucial robotics-related tasks, such as multi-robot navigation and obstacle avoidance, that have been conventionally approached with the classical non-learnable methods (e.g., heuristic search) is currently suggested to be solved by the learning-based or hybrid methods. Still, in this domain, it is hard, not to say impossible, to conduct a fair comparison between classical, learning-based, and hybrid approaches due to the lack of a unified framework that supports both learning and evaluation. To this end, we introduce POGEMA, a set of comprehensive tools that includes a fast environment for learning, a generator of problem instances, the collection of pre-defined ones, a visualization toolkit, and a benchmarking tool that allows automated evaluation. We introduce and specify an evaluation protocol defining a range of domain-related metrics computed on the basics of the primary evaluation indicators (such as success rate and path length), allowing a fair multi-fold comparison. The results of such a comparison, which involves a variety of state-of-the-art MARL, search-based, and hybrid methods, are presented., Comment: 27 pages, 9 figures

Handling vast amounts of data is crucial in today's world. The growth of high-performance computing has created a need for parallelization, particularly in the area of machine learning algorithms such as ANN (Approximate Nearest Neighbors). To improve the speed of these algorithms, it is important to optimize them for specific processor architectures. RISC-V (Reduced Instruction Set Computer Five) is one of the modern processor architectures, which features a vector instruction set called RVV (RISC-V Vector Extension). In machine learning algorithms, vector extensions are widely utilized to improve the processing of voluminous data. This study examines the effectiveness of applying RVV to commonly used ANN algorithms. The algorithms were adapted for RISC-V and optimized using RVV after identifying the primary bottlenecks. Additionally, we developed a theoretical model of a parameterized vector block and identified the best on average configuration that demonstrates the highest theoretical performance of the studied ANN algorithms when the other CPU parameters are fixed.

When silica-filled rubber compounds are deformed, structural modifications in the material's bulk lead to irreversible damage, the most significant of which is cavitation appearing within the interfaces of interconnected polymer and filler networks. This work introduces a new method to analyze cavitation in industrial-grade rubbers based on Ultra Small-Angle X-ray Scattering. This method employs a specially designed multi-sample stretching device for high-throughput measurements with statistical relevance. The proposed data reduction approach allows for early detection and quantification of cavitation while providing at the same time information on the hierarchical filler structures at length scales ranging from the primary particle size to large silica agglomerates over four orders of magnitude. To validate the method, the scattering of SSBR rubber compounds filled with highly dispersible silica at different ratios was measured under quasi-static strain. The strain was applied in incremental steps up to a maximum achievable elongation or breakage of the sample. From the measurements performed in multiple repetitions, it was found that the minimum strain necessary for cavity formation and the size evolution of the cavities with increasing strain are comparable between these samples. The sample with the highest polymer content showed the lowest rate of cavity formation and higher durability of silica structures. The structural stability of the compounds was determined by the evolution of the filler hierarchical structures, obtained by fitting data across the available strain range., Comment: 17 pages with 2 tables and 13 figures

The biomechanical properties of cells and tissues play an important role in our fundamental understanding of the structures and functions of biological systems at both the cellular and subcellular levels. Recently, Brillouin microscopy, which offers a label-free spectroscopic means of assessing viscoelastic properties in vivo, has emerged as a powerful way to interrogate those properties on a microscopic level in living tissues. However, susceptibility to photo-damage and photo-bleaching, particularly when high-intensity laser beams are used to induce Brillouin scattering, poses a significant challenge. This article introduces a transformative approach designed to mitigate photo-damage in biological and biomedical studies, enabling non-destructive, label-free assessments of mechanical properties in live biological samples. By leveraging quantum-light-enhanced stimulated Brillouin scattering (SBS) imaging contrast, the signal-to-noise ratio is significantly elevated, thereby increasing sample viability and extending interrogation times without compromising the integrity of living samples. The tangible impact of this novel methodology is evidenced by a notable three-fold increase in sample viability observed after subjecting the samples to three hours of continuous squeezed-light illumination, surpassing the traditional coherent light-based approaches. The quantum-enhanced SBS imaging holds promise across diverse fields, such as cancer biology and neuroscience where preserving sample vitality is of paramount significance. By mitigating concerns regarding photo-damage and photo-bleaching associated with high-intensity lasers, this technological breakthrough expands our horizons for exploring the mechanical properties of live biological systems, paving the way for a new era of research and clinical applications., Comment: 14 pages, 7 figures

Quasi-parametric amplification (QPA) is a nonlinear interaction in which the idler wave is depleted through some loss mechanism. QPA plays an important role in signal amplification in ultrafast photonics and quantum light generation. The QPA process has a number of features characterized by the non-Hermitian parity-time ($\mathcal{PT}$) symmetry. In this report, we explore new interaction regimes and uncover multiple $\mathcal{PT}$-symmetry phase transitions in such QPA process where transitions are particularly sensitive to external parameters. In particular, we demonstrate the feasibility of detection of $10^{-11}$ inhomogeneities of the doped absorber, which is order of magnitude more sensitive than similar measurements performed in a linear absorption regime. In doing so, we reveal a family of $\mathcal{PT}$-symmetry phase transitions appearing in the QPA process and provide a novel nonlinear optical sensing mechanism for precise optical measurements., Comment: 17 pages, 6 figures

Multilayer van der Waals (vdW) heterostructures have become an important platform in which to study novel fundamental effects emerging at the nanoscale. Standard nanopatterning techniques relying on electron-beam lithography and reactive ion etching, widely applied to pattern such heterostructures, however, impose some limitations on the edge accuracy and resolution, as revealed through numerous experiments with vdW quantum dots and point contacts. Here we present an alternative approach for electrode-free nanopatterning of thick multilayer vdW heterostructures based on atomic force microscopy (AFM). By applying an AC voltage of a relatively small frequency (1-10 kHz) between the sharp platinum tip and the substrate, we realize high-resolution ($\lesssim 100$ nm) etching of thick multicomponent heterostructures if the latter are deposited onto graphite slabs. Importantly, unlike more conventional electrode-free local anodic oxidation, our method does not require a special environment with excess humidity, can be applied at ambient conditions, and enables the patterning of multilayer heterostructures composed of graphene, graphite, hexagonal boron nitride (hBN), NbSe$_{2}$, WSe$_{2}$, and more., Comment: 6 pages; 3 figures

We describe a gyroscope that measures rotation based on the effects of the rotation on the polarization of light. Rotation induces a differential phase shift in the propagation of left and right circularly polarized light and this phase shift can be measured in suitably designed interferometric setups. The signal in this setup is independent of the frequency of light, unlike various sources of noise such as vibrations, which cause phase shifts that depend on the frequency. Such vibrations are the practical limit on the sensitivity of conventional Sagnac-style optical interferometers that are typically used as gyroscopes. In the proposed setup, one can potentially mitigate this source of noise by simultaneously using two (or more) sources of light that have different frequencies. The signal in this setup scales with the total storage time of the light. Due to its frequency independence, it is thus most optimal to measure the signal using superconducting RF systems where the high finesse of the available cavities enables considerably longer storage times than is possible in an optical setup., Comment: 15 pages, 1 figure

Variational inequalities represent a broad class of problems, including minimization and min-max problems, commonly found in machine learning. Existing second-order and high-order methods for variational inequalities require precise computation of derivatives, often resulting in prohibitively high iteration costs. In this work, we study the impact of Jacobian inaccuracy on second-order methods. For the smooth and monotone case, we establish a lower bound with explicit dependence on the level of Jacobian inaccuracy and propose an optimal algorithm for this key setting. When derivatives are exact, our method converges at the same rate as exact optimal second-order methods. To reduce the cost of solving the auxiliary problem, which arises in all high-order methods with global convergence, we introduce several Quasi-Newton approximations. Our method with Quasi-Newton updates achieves a global sublinear convergence rate. We extend our approach with a tensor generalization for inexact high-order derivatives and support the theory with experiments.