Search

Semantic segmentation on LiDAR imaging is increasingly gaining attention, as it can provide useful knowledge for perception systems and potential for autonomous driving. However, collecting and labeling real LiDAR data is an expensive and time-consuming task. While datasets such as SemanticKITTI have been manually collected and labeled, the introduction of simulation tools such as CARLA, has enabled the creation of synthetic datasets on demand. In this work, we present a modified CARLA simulator designed with LiDAR semantic segmentation in mind, with new classes, more consistent object labeling with their counterparts from real datasets such as SemanticKITTI, and the possibility to adjust the object class distribution. Using this tool, we have generated SynthmanticLiDAR, a synthetic dataset for semantic segmentation on LiDAR imaging, designed to be similar to SemanticKITTI, and we evaluate its contribution to the training process of different semantic segmentation algorithms by using a naive transfer learning approach. Our results show that incorporating SynthmanticLiDAR into the training process improves the overall performance of tested algorithms, proving the usefulness of our dataset, and therefore, our adapted CARLA simulator. The dataset and simulator are available in https://github.com/vpulab/SynthmanticLiDAR., Comment: 2024 IEEE International Conference on Image Processing (ICIP)

Semantic segmentation is a computer vision task where classification is performed at a pixel level. Due to this, the process of labeling images for semantic segmentation is time-consuming and expensive. To mitigate this cost there has been a surge in the use of synthetically generated data -- usually created using simulators or videogames -- which, in combination with domain adaptation methods, can effectively learn how to segment real data. Still, these datasets have a particular limitation: due to their closed-set nature, it is not possible to include novel classes without modifying the tool used to generate them, which is often not public. Concurrently, generative models have made remarkable progress, particularly with the introduction of diffusion models, enabling the creation of high-quality images from text prompts without additional supervision. In this work, we propose an unsupervised pipeline that leverages Stable Diffusion and Segment Anything Module to generate class examples with an associated segmentation mask, and a method to integrate generated cutouts for novel classes in semantic segmentation datasets, all with minimal user input. Our approach aims to improve the performance of unsupervised domain adaptation methods by introducing novel samples into the training data without modifications to the underlying algorithms. With our methods, we show how models can not only effectively learn how to segment novel classes, with an average performance of 51% IoU, but also reduce errors for other, already existing classes, reaching a higher performance level overall.

This paper introduces a novel synthetic dataset that captures urban scenes under a variety of weather conditions, providing pixel-perfect, ground-truth-aligned images to facilitate effective feature alignment across domains. Additionally, we propose a method for domain adaptation and generalization that takes advantage of the multiple versions of each scene, enforcing feature consistency across different weather scenarios. Our experimental results demonstrate the impact of our dataset in improving performance across several alignment metrics, addressing key challenges in domain adaptation and generalization for segmentation tasks. This research also explores critical aspects of synthetic data generation, such as optimizing the balance between the volume and variability of generated images to enhance segmentation performance. Ultimately, this work sets forth a new paradigm for synthetic data generation and domain adaptation.

Quantum tokens envision to store unclonable authentication keys in quantum states that are issued by a bank for example. In contrast to quantum communication, the information is not transmitted, but rather used for personal authentication in a physical device. Still, its experimental realization faces many technical challenges. In this work, we propose an ensemble-based quantum-token protocol, making these applications technologically less-demanding. A simple and minimal model is developed to describe the quantum token hardware, while the protocol is fully benchmarked and compared on five different IBM quantum processors. First, the uncertainties of the hardware are characterized, from which the main quality parameters that describe the token can be extracted. Following that, the fraction of qubits which the bank prepares and measures successfully is benchmarked. These fractions are then compared with the values obtained from an attacker who attempts to read the bank token and prepare a forged key. From which we experimentally demonstrate an acceptance probability of 0.057 for a forged token, in contrast to 0.999 for the bank's own tokens. These values can be further optimized by increasing the number of tokens in the device. Finally, we show that minor improvements in the hardware quality lead to significant increases in the protocol security, denoting a great potential of the protocol to scale with the ongoing quantum hardware evolution. We provide an open source tool with graphical user interface to benchmark the protocol with custom ensemble based qubits. This work demonstrates the overall security of the protocol within a hardware-agnostic framework, further confirming the interoperability of the protocol in arbitrary quantum systems and thus paving the way for future applications with different qubits., Comment: 14 pages, 10 figures, associated Github repository https://github.com/lucas-tsunaki/quantum-token

Due to the difficulty of replicating the real conditions during training, supervised algorithms for spacecraft pose estimation experience a drop in performance when trained on synthetic data and applied to real operational data. To address this issue, we propose a test-time adaptation approach that leverages the temporal redundancy between images acquired during close proximity operations. Our approach involves extracting features from sequential spacecraft images, estimating their poses, and then using this information to synthesise a reconstructed view. We establish a self-supervised learning objective by comparing the synthesised view with the actual one. During training, we supervise both pose estimation and image synthesis, while at test-time, we optimise the self-supervised objective. Additionally, we introduce a regularisation loss to prevent solutions that are not consistent with the keypoint structure of the spacecraft. Our code is available at: https://github.com/JotaBravo/spacecraft-tta., Comment: Preprint

Highly non-equilibrium conditions in femtosecond-laser excited solids cause a variety of ultrafast phenomena that are not accessible by thermal conditions, like sub-picosecond solid-to-liquid or solid-to-solid phase transitions. In recent years the microscopic pathways of various laser-induced crystal rearrangements could be identified and led to novel applications and/or improvements in optoelectronics, photonics, and nanotechnology. However, it remains unclear what effect a femtosecond-laser excitation has on ionic impurities within an altered crystal environment, in particular on the atomic mobility. Here, we performed ab-initio molecular dynamics (AIMD) simulations on laser-excited black silicon, a promising material for high-efficient solar cells, using the Code for Highly excIted Valence Electron Systems (CHIVES). By computing time-dependent Bragg peak intensities for doping densities of 0.16% and 2.31% we could identify the overall weakening of the crystal environment with increasing impurity density. The analysis of Si-S bond angles and lengths after different excitation densities, as well as computing interatomic forces allowed to identify a change in ion mobility with increasing impurity density and excitation strength. Our results indicate the importance of impurity concentrations for ionic mobility in laser-excited black silicon and could give significant insight for semiconductor device optimization and materials science advancement., Comment: 12 pages, 9 figures

Acoustoplasmonic resonators, such as nanobars and crosses, are efficient acousto-optical transducers. The excitation of mechanical modes in these structures strongly depends on the spatial profile of the eigenmodes of the resonator. Using a system of two identical gold elongated bars placed on a silicon dioxide substrate, we examine how breaking mirror symmetries affects the optical and acoustic properties to provide insights in the design of acoustoplasmonic metasurfaces for nonsymmetric acousto-optical transducers. Our findings show that, the absence of mirror symmetries affects differently the optical and nanomechanical response. Broken mirror symmetries not only couple nanomechanical modes existing in single bars, but introduces new torsional resonant modes.

We introduce a framework for simulating, on an $(n+1)$-qubit quantum computer, the action of a Gaussian Bosonic (GB) circuit on a state over $2^n$ modes. Specifically, we encode the initial bosonic state's expectation values over quadrature operators (and their covariance matrix) as an input qubit-state. This is then evolved by a quantum circuit that effectively implements the symplectic propagators induced by the GB gates. We find families of GB circuits and initial states leading to efficient quantum simulations. For this purpose, we introduce a dictionary that maps between GB and qubit gates such that particle- (non-particle-) preserving GB gates lead to real (imaginary) time evolutions at the qubit level. For the special case of particle-preserving circuits, we present a BQP-complete GB decision problem, indicating that GB evolutions of Gaussian states on exponentially many modes are as powerful as universal quantum computers. We also perform numerical simulations of an interferometer on $\sim8$ billion modes, illustrating the power of our framework., Comment: 5+14 pages, 3 figures, 1 table

Non-conventional resonances, both acoustic and photonic, are found in metallic particles with a toroidal nanopropeller geometry that is generated by sweeping a three-lobed 2D-shape along a spiral with twisting angle, ${\alpha}$. For both optical and acoustic cases, spectral location of resonances experiences a red-shift as a function of ${\alpha}$. We demonstrate that the optical case can be understood as a natural evolution of resonances as the spiral length of the toroidal nanopropeller increases with ${\alpha}$, implying a huge helicity dependent absorption cross section. In the case of acoustic response, two red-shifting breathing modes are identified. Additionally, even small ${\alpha}$ allows the appearance of new low-frequency resonances, whose spectral dispersion depends on a competition between length of the generative spiral and the pitch of the toroidal nanopropeller., Comment: 9 pages 7 figures

The scarcity of data acquired under actual space operational conditions poses a significant challenge for developing learning-based visual navigation algorithms crucial for autonomous spacecraft navigation. This data shortage is primarily due to the prohibitive costs and inherent complexities of space operations. While existing datasets, predominantly relying on computer-simulated data, have partially addressed this gap, they present notable limitations. Firstly, these datasets often utilize proprietary image generation tools, restricting the evaluation of navigation methods in novel, unseen scenarios. Secondly, they provide limited ground-truth data, typically focusing solely on the spacecraft's translation and rotation relative to the camera. To address these limitations, we present SPIN (SPacecraft Imagery for Navigation), an open-source spacecraft image generation tool designed to support a wide range of visual navigation scenarios in space, with a particular focus on relative navigation tasks. SPIN provides multiple modalities of ground-truth data and allows researchers to employ custom 3D models of satellites, define specific camera-relative poses, and adjust settings such as camera parameters or environmental illumination conditions. We also propose a method for exploiting our tool as a data augmentation module. We validate our tool on the spacecraft pose estimation task by training with a SPIN-generated replica of SPEED+, reaching a 47% average error reduction on SPEED+ testbed data (that simulates realistic space conditions), further reducing it to a 60% error reduction when using SPIN as a data augmentation method. Both the SPIN tool (and source code) and our SPIN-generated version of SPEED+ will be publicly released upon paper acceptance on GitHub. https://github.com/vpulab/SPIN

Parametrized and random unitary (or orthogonal) $n$-qubit circuits play a central role in quantum information. As such, one could naturally assume that circuits implementing symplectic transformation would attract similar attention. However, this is not the case, as $\mathbb{SP}(d/2)$ -- the group of $d\times d$ unitary symplectic matrices -- has thus far been overlooked. In this work, we aim at starting to right this wrong. We begin by presenting a universal set of generators $\mathcal{G}$ for the symplectic algebra $i\mathfrak{sp}(d/2)$, consisting of one- and two-qubit Pauli operators acting on neighboring sites in a one-dimensional lattice. Here, we uncover two critical differences between such set, and equivalent ones for unitary and orthogonal circuits. Namely, we find that the operators in $\mathcal{G}$ cannot generate arbitrary local symplectic unitaries and that they are not translationally invariant. We then review the Schur-Weyl duality between the symplectic group and the Brauer algebra, and use tools from Weingarten calculus to prove that Pauli measurements at the output of Haar random symplectic circuits can converge to Gaussian processes. As a by-product, such analysis provides us with concentration bounds for Pauli measurements in circuits that form $t$-designs over $\mathbb{SP}(d/2)$. To finish, we present tensor-network tools to analyze shallow random symplectic circuits, and we use these to numerically show that computational-basis measurements anti-concentrate at logarithmic depth., Comment: 13+7 pages, 8 figures

With a single circulating vector-borne virus, the basic reproduction number incorporates contributions from tick-to-tick (co-feeding), tick-to-host and host-to-tick transmission routes. With two different circulating vector-borne viral strains, resident and invasive, and under the assumption that co-feeding is the only transmission route in a tick population, the invasion reproduction number depends on whether the model system of ordinary differential equations possesses the property of neutrality. We show that a simple model, with two populations of ticks infected with one strain, resident or invasive, and one population of co-infected ticks, does not have Alizon's neutrality property. We present model alternatives that are capable of representing the invasion potential of a novel strain by including populations of ticks dually infected with the same strain. The invasion reproduction number is analysed with the next-generation method and via numerical simulations., Comment: 37 pages and 4 figures

The ultrafast dynamics of ions in solids following intense femtosecond laser excitation is governed by two fundamentally distinct yet interplaying effects. On one hand, the significant generation of hot electron-hole pairs by the light pulse alters the strength and nature of interatomic bonding, resulting in nonthermal ionic motion. On the other hand, incoherent electron-ion collisions drives an equilibration of the electrons and the ions, which reach a common temperature on a picosecond timescale. This letter introduces, for the first time, a unified theory that comprehensively accounts for both phenomena. Our approach is versatile, applicable to both ab-initio and large-scale molecular dynamics simulations, and leads to a generalization of the two-temperature model-molecular-dynamics equations of motion. To illustrate the effectiveness of our method, we apply it to describe the laser excitation of silicon thin films. Our simulations reproduce the time evolution of the Bragg peaks in excellent agreement with the experiments., Comment: 12 pages, 2 figures, 20 equations

We introduce a variational quantum solver for combinatorial optimizations over $m=\mathcal{O}(n^k)$ binary variables using only $n$ qubits, with tunable $k>1$. The number of parameters and circuit depth display mild linear and sublinear scalings in $m$, respectively. Moreover, we analytically prove that the specific qubit-efficient encoding brings in a super-polynomial mitigation of barren plateaus as a built-in feature. This leads to unprecedented quantum-solver performances. For $m=7000$, numerical simulations produce solutions competitive in quality with state-of-the-art classical solvers. In turn, for $m=2000$, an experiment with $n=17$ trapped-ion qubits featured MaxCut approximation ratios estimated to be beyond the hardness threshold $0.941$. To our knowledge, this is the highest quality attained experimentally on such sizes. Our findings offer a novel heuristics for quantum-inspired solvers as well as a promising route towards solving commercially-relevant problems on near term quantum devices.

Determining the size distribution of asteroids is key for understanding the collisional history and evolution of the inner Solar System. We aim at improving our knowledge on the size distribution of small asteroids in the Main Belt by determining the parallaxes of newly detected asteroids in the Hubble Space Telescope (HST) Archive and hence their absolute magnitudes and sizes. Asteroids appear as curved trails in HST images due to the parallax induced by the fast orbital motion of the spacecraft. The parallax effect can be computed to obtain the distance to the asteroids by fitting simulated trajectories to the observed trails. Using distance, we can obtain the object's absolute magnitude and size estimation assuming an albedo value, along with some boundaries for its orbital parameters. In this work we analyse a set of 632 serendipitously imaged asteroids found in the ESA HST Archive. An object-detection machine learning algorithm was used to perform this task during previous work. Our raw data consists of 1,031 asteroids trails from unknown objects (not matching any entries in the MPC database). We also found 670 trails from known objects (objects featuring matching entries in the MPC). After an accuracy assessment and filtering process, our analysed HST set consists of 454 unknown objects and 178 known objects. We obtain a sample dominated by potential Main Belt objects featuring absolute magnitudes (H) mostly between 15 and 22 mag. The absolute magnitude cumulative distribution confirms the previously reported slope change for 15 < H < 18, from 0.56 to 0.26, maintained in our case down to absolute magnitudes around H = 20, hence expanding the previous results by approximately two magnitudes. HST archival observations can be used as an asteroid survey since the telescope pointings are statistically randomly oriented in the sky and they cover long periods of time., Comment: Accepted for publication in A&A. 10 pages, 2 tables, 17 figures

A large amount of effort has recently been put into understanding the barren plateau phenomenon. In this perspective article, we face the increasingly loud elephant in the room and ask a question that has been hinted at by many but not explicitly addressed: Can the structure that allows one to avoid barren plateaus also be leveraged to efficiently simulate the loss classically? We present strong evidence that commonly used models with provable absence of barren plateaus are also classically simulable, provided that one can collect some classical data from quantum devices during an initial data acquisition phase. This follows from the observation that barren plateaus result from a curse of dimensionality, and that current approaches for solving them end up encoding the problem into some small, classically simulable, subspaces. Thus, while stressing quantum computers can be essential for collecting data, our analysis sheds serious doubt on the non-classicality of the information processing capabilities of parametrized quantum circuits for barren plateau-free landscapes. We end by discussing caveats in our arguments, the role of smart initializations and the possibility of provably superpolynomial, or simply practical, advantages from running parametrized quantum circuits., Comment: 14+15 pages, 5 figures, 2 tables, minor corrections added

Barren plateaus have emerged as a pivotal challenge for variational quantum computing. Our understanding of this phenomenon underwent a transformative shift with the recent introduction of a Lie algebraic theory capable of explaining most sources of barren plateaus. However, this theory requires either initial states or observables that lie in the circuit's Lie algebra. Focusing on parametrized matchgate circuits, in this work we are able to go beyond this assumption and provide an exact formula for the loss function variance that is valid for arbitrary input states and measurements. Our results reveal that new phenomena emerge when the Lie algebra constraint is relaxed. For instance, we find that the variance does not necessarily vanish inversely with the Lie algebra's dimension. Instead, this measure of expressiveness is replaced by a generalized expressiveness quantity: The dimension of the Lie group modules. By characterizing the operators in these modules as products of Majorana operators, we can introduce a precise notion of generalized globality and show that measuring generalized-global operators leads to barren plateaus. Our work also provides operational meaning to the generalized entanglement as we connect it with known fermionic entanglement measures, and show that it satisfies a monogamy relation. Finally, while parameterized matchgate circuits are not efficiently simulable in general, our results suggest that the structure allowing for trainability may also lead to classical simulability., Comment: 5+26 pages, 2+1 figures

The use of rare-earth-based permanent magnets is one of the critical points for the development of the current technology. On the one hand, industry of the rare-earths is highly polluting due to the negative environmental impact of their extraction and, on the other hand, the sector is potentially dependent on China. Therefore, investigation is required both in the development of rare-earth-free permanent magnets and in sintering processes that enable their greener fabrication with attractive magnetic properties at a more competitive price. This work presents the use of a cold sintering process (CSP) followed by a post-annealing at 1100 {\deg}C as a new way to sinter composite permanent magnets based on strontium ferrite (SFO). Composites that incorporate a percentage < 10% of an additional magnetic phase have been prepared and the morphological, structural and magnetic properties have been evaluated after each stage of the process. CSP induces a phase transformation of SFO in the composites, which is partially recovered by the post-thermal treatment improving the relative density to 92% and the magnetic response of the final magnets with a coercivity of up to 3.0 kOe. Control of the magnetic properties is possible through the composition and the grain size in the sintered magnets. These attractive results show the potential of the sintering approach as an alternative to develop modern rare-earth-free composite permanent magnets.

Owing to their multiple applications, lithium ferrites are relevant materials for several emerging technologies. For instance, LiFeO2 has been spotted as an alternative cathode material in Li-ion batteries, while LiFe5O8 is the lowest damping ferrite, holding promise in the field of spintronics. The Li-content in lithium ferrites has been shown to greatly affect the physical properties, and in turn, the performance of functional devices based on these materials. Despite this, lithium content is rarely accurately quantified, as a result of the low number of electrons in Li hindering its identification by means of routine materials characterization methods. In the present work, magnetic lithium ferrite powders with Li:Fe ratios of 1:1, 1:3 and 1:5 have been synthesized, successfully obtaining phase-pure materials (LiFeO2 and LiFe5O8), as well as a controlled mixture of both phases. The powders have been compacted and subsequently sintered by thermal treatment (Tmax = 1100 {\deg}C) to fabricate dense pellets which preserve the original Li:Fe ratios. Li-content on both powders and pellets has been determined by two independent methods: (i) Rutherford backscattering spectroscopy combined with nuclear reaction analysis and (ii) Rietveld analysis of powder X-ray diffraction data. With good agreement between both techniques, it has been confirmed that the Li:Fe ratios employed in the synthesis are maintained in the sintered ceramics. The same conclusion is drawn from spatially-resolved confocal Raman microscopy experiments on regions of a few microns. Field emission scanning electron microscopy has evidenced the substantial grain growth taking place during the sintering process - mean particle sizes rise from about 600 nm in the powders up to 3.8(6) um for dense LiFeO2 and 10(2) um for LiFe5O8 ceramics.

In this work we have investigated the effect of the solvent during the processing of SrFe12O19 platelet-based permanent magnets by cold sintering process (CSP) plus a post-thermal treatment. Several organic solvents: glacial acetic acid, oleic acid and oleylamine have been analyzed, optimizing the CSP temperatures at 190-270 {\deg}C, under pressures of 375-670 MPa and 6-50 wt% of solvent. Modifications in the morphological and structural properties are identified depending on the solvent, which impacts on the magnetic response. Independently of the solvent, the mechanical integrity of ferrite magnets obtained by CSP is improved by a post-annealing at 1100 {\deg}C for 2 h, resulting in relative densities around 92 % with an average grain size of 1 {\mu}m and a fraction of SrFe12O19 phase > 91 %. Hc > 2.1 kOe and MS of 73 emu/g are obtained in the final sintered ceramic magnets, exhibiting the highest HC value of 2.8 kOe for the magnet sintered using glacial acetic acid.

The incessant technological pursuit towards a more sustainable and green future depends strongly on permanent magnets. At present, their use is widespread, making it imperative to develop new processing methods that generate highly competitive magnetic properties reducing the fabrication temperatures and costs. Herein, a novel strategy for developing dense sintered magnets based on Sr-hexaferrites with upper functional characteristics is presented. An innovative cold sintering approach using glacial acetic acid as novelty, followed by a post-annealing at 1100 {\deg}C, achieves a densification of the ceramic magnets of 92% with respect to the theoretical density and allows controlling the particle growth. After the cold sintering process, a fraction of amorphous SrO is identified, in addition to a partial transformation to {\alpha}-Fe2O3 as secondary crystalline phase. 46 wt% of SrFe12O19 remains, which is mostly recuperated after the post-thermal treatment. These findings do not significantly modify the final structure of ferrite magnets, neither at short- nor long-range order. The innovative process has a positive impact on the magnetic properties, yielding competitive ferrite magnets at lower sintering temperatures with an energy efficiency of at least 25%, which opens up a new horizon in the field of rare-earth free permanent magnets and new possibilities in other applications.

Anisotropic hybrid nanostructures stand out as promising therapeutic agents in photothermal conversion-based treatments. Accordingly, understanding local heat generation mediated by light-to-heat conversion of absorbing multicomponent nanoparticles at the single particle level have both forthwith become a subject of broad and current interest. Nonetheless, evaluating reliable temperature profiles around a single trapped nanoparticle is challenging from all experimental, computational, and fundamental viewpoints. Committed to filling this gap, the heat generation of an anisotropic hybrid nanostructure is explored by means of two different experimental approaches from which the local temperature is measured in a direct or indirect way, all in the context of the hot Brownian motion theory. The results were compared with analytical results supported by the numerical computation of the wavelength-dependent absorption efficiencies in the discrete dipole approximation for scattering calculations, which has been here extended to inhomogeneous nanostructures. Overall, we provide a consistent and comprehensive view of the heat generation in optical traps of highly absorbing particles under the viewpoint of the hot Brownian motion theory.

An intense femtosecond-laser excitation of a solid induces highly nonthermal conditions. In materials like silicon, laser-induced bond-softening leads to a highly incoherent ionic motion and eventually nonthermal melting. But is this outcome an inevitable consequence, or can it be controlled? Here, we performed ab initio molecular dynamics simulations of crystalline silicon after timed multiple femtosecond-laser pulse excitations with fluence above the nonthermal melting threshold. Our results demonstrate an excitation mechanism that pauses nonthermal melting and creates a metastable state instead, with an electronic structure similar to the ground state. This mechanism can be generalized to other materials, potentially enabling structural and/or electronic transitions to metastable phases in the high-excitation regime. In addition, our approach could be used to switch off nonthermal contributions in experiments, allowing reliable electron-phonon coupling constants to be obtained more easily., Comment: Letter plus Suppl. Mat

Ordinary differential equation (ODE) models used in mathematical epidemiology assume explicitly or implicitly large populations. For the study of infections in a hospital this is an extremely restrictive assumption as typically a hospital ward has a few dozen, or even fewer, patients. This work reframes a well-known model used in the study of the spread of antibiotic-resistant bacteria in hospitals, to consider the pathogen transmission dynamics in small populations. In this vein, this paper proposes a Markov chain model to describe the spread of a single bacterial species in a hospital ward where patients may be free of bacteria or may carry bacterial strains that are either sensitive or resistant to antimicrobial agents. We determine the probability law of the \emph{exact} reproduction number ${\cal R}_{exact,0}$, which is here defined as the random number of secondary infections generated by those patients who are accommodated in a predetermined bed before a patient who is free of bacteria is accommodated in this bed for the first time. Specifically, we decompose the exact reproduction number ${\cal R}_{exact,0}$ into two contributions allowing us to distinguish between infections due to the sensitive and the resistant bacterial strains. Our methodology is mainly based on structured Markov chains and the use of related matrix-analytic methods. This guarantees the compatibility of the new, finite-population model, with large population models present in the literature and takes full advantage, in its mathematical analysis, of the intrinsic stochasticity., Comment: 30 pages, 9 figures

It is well known that artificial neural networks initialized from independent and identically distributed priors converge to Gaussian processes in the limit of a large number of neurons per hidden layer. In this work we prove an analogous result for Quantum Neural Networks (QNNs). Namely, we show that the outputs of certain models based on Haar random unitary or orthogonal deep QNNs converge to Gaussian processes in the limit of large Hilbert space dimension $d$. The derivation of this result is more nuanced than in the classical case due to the role played by the input states, the measurement observable, and the fact that the entries of unitary matrices are not independent. Then, we show that the efficiency of predicting measurements at the output of a QNN using Gaussian process regression depends on the observable's bodyness. Furthermore, our theorems imply that the concentration of measure phenomenon in Haar random QNNs is worse than previously thought, as we prove that expectation values and gradients concentrate as $\mathcal{O}\left(\frac{1}{e^d \sqrt{d}}\right)$. Finally, we discuss how our results improve our understanding of concentration in $t$-designs., Comment: 14+37 pages, 4+6 figures

All possible phenotypes are not equally accessible to evolving populations. In fact, only phenotypes of large size, i.e. those resulting from many different genotypes, are found in populations of sequences, presumably because they are easier to discover and maintain. Genotypes that map to these phenotypes usually form mostly connected genotype networks that percolate the space of sequences, thus guaranteeing access to a large set of alternative phenotypes. Within a given environment, where specific phenotypic traits become relevant for adaptation, the replicative ability of a phenotype and its overall fitness (in competition experiments with alternative phenotypes) can be estimated. Two primary questions arise: how do phenotype size, reproductive capability and topology of the genotype network affect the fitness of a phenotype? And, assuming that evolution is only able to access large phenotypes, what is the range of unattainable fitness values? In order to address these questions, we quantify the adaptive advantage of phenotypes of varying size and spectral radius in a two-peak landscape. We derive analytical relationships between the three variables (size, topology, and replicative ability) which are then tested through analysis of genotype-phenotype maps and simulations of population dynamics on such maps. Finally, we analytically show that the fraction of attainable phenotypes decreases with the length of the genotype, though its absolute number increases. The fact that most phenotypes are not visible to evolution very likely forbids the attainment of the highest peak in the landscape. Nevertheless, our results indicate that the relative fitness loss due to this limited accessibility is largely inconsequential for adaptation., Comment: 25 pages, 10 figures, uses iopart.cls, iopart10.clo, iopart12.clo, iopams.sty, setstack.sty

Common models of circadian rhythms are constructed as compartmental reactions of well mixed biochemicals involving a negative-feedback loop containing several intermediate reaction steps in order to enable oscillations. Spatial transport of reactants is mimicked as an extra compartmental reaction step. In this letter, we show that a single activation-repression biochemical reaction pair is enough to produce sustained oscillations, if the sites of both reactions are spatially separated and molecular transport is mediated by diffusion. Our proposed scenario is the simplest possible one in terms of the participating chemical reactions and provides a conceptual basis for understanding biological oscillations and triggering in-vitro assays aimed at constructing minimal clocks., Comment: 13 pages, 4 figures (main), 16 pages, 4 figures (supplemental)

Microgels are soft microparticles that often exhibit thermoresponsiveness and feature a transformation at a critical temperature, referred to as the volume phase transition temperature. The question of whether this transformation occurs as a smooth or as a discontinuous one is still a matter of debate. This question can be addressed by studying individual microgels trapped in optical tweezers. For this aim, composite particles were obtained by decorating pNIPAM microgels with iron oxide nanocubes. These composites become self-heating when illuminated by the infrared trapping laser, featuring Hot Brownian Motion within the trap. Above a certain laser power, a single decorated microgel features a volume phase transition that is discontinuous, while the usual continuous sigmoidal-like dependence is recovered after averaging over different microgels. The collective sigmoidal behavior enables the application of a power-to-temperature calibration and provides the effective drag coefficient of the self-heating microgels, thus establishing these composite particles as potential micro-thermometers and micro-heaters. Moreover, the self-heating microgels also exhibit an unexpected and intriguing bistability behavior above the critical temperature, probably due to partial collapses of the microgel. These results set the stage for further studies and the development of applications based on the Hot Brownian Motion of soft particles.

Background: Several studies indicate that the anterior visual pathway provides information about the dynamics of axonal degeneration in Multiple Sclerosis (MS). Current research in the field is focused on the quest for the most discriminative features among patients and controls and the development of machine learning models that yield computer-aided solutions widely usable in clinical practice. However, most studies are conducted with small samples and the models are used as black boxes. Clinicians should not trust machine learning decisions unless they come with comprehensive and easily understandable explanations. Materials and methods: A total of 216 eyes from 111 healthy controls and 100 eyes from 59 patients with relapsing-remitting MS were enrolled. The feature set was obtained from the thickness of the ganglion cell layer (GCL) and the retinal nerve fiber layer (RNFL). Measurements were acquired by the novel Posterior Pole protocol from Spectralis Optical Coherence Tomography (OCT) device. We compared two black-box methods (gradient boosting and random forests) with a glass-box method (explainable boosting machine). Explainability was studied using SHAP for the black-box methods and the scores of the glass-box method. Results: The best-performing models were obtained for the GCL layer. Explainability pointed out to the temporal location of the GCL layer that is usually broken or thinning in MS and the relationship between low thickness values and high probability of MS, which is coherent with clinical knowledge. Conclusions: The insights on how to use explainability shown in this work represent a first important step toward a trustworthy computer-aided solution for the diagnosis of MS with OCT.

Overparametrization is one of the most surprising and notorious phenomena in machine learning. Recently, there have been several efforts to study if, and how, Quantum Neural Networks (QNNs) acting in the absence of hardware noise can be overparametrized. In particular, it has been proposed that a QNN can be defined as overparametrized if it has enough parameters to explore all available directions in state space. That is, if the rank of the Quantum Fisher Information Matrix (QFIM) for the QNN's output state is saturated. Here, we explore how the presence of noise affects the overparametrization phenomenon. Our results show that noise can "turn on" previously-zero eigenvalues of the QFIM. This enables the parametrized state to explore directions that were otherwise inaccessible, thus potentially turning an overparametrized QNN into an underparametrized one. For small noise levels, the QNN is quasi-overparametrized, as large eigenvalues coexists with small ones. Then, we prove that as the magnitude of noise increases all the eigenvalues of the QFIM become exponentially suppressed, indicating that the state becomes insensitive to any change in the parameters. As such, there is a pull-and-tug effect where noise can enable new directions, but also suppress the sensitivity to parameter updates. Finally, our results imply that current QNN capacity measures are ill-defined when hardware noise is present., Comment: 14 + 6 pages, 11 figures

Spacecraft pose estimation is a key task to enable space missions in which two spacecrafts must navigate around each other. Current state-of-the-art algorithms for pose estimation employ data-driven techniques. However, there is an absence of real training data for spacecraft imaged in space conditions due to the costs and difficulties associated with the space environment. This has motivated the introduction of 3D data simulators, solving the issue of data availability but introducing a large gap between the training (source) and test (target) domains. We explore a method that incorporates 3D structure into the spacecraft pose estimation pipeline to provide robustness to intensity domain shift and we present an algorithm for unsupervised domain adaptation with robust pseudo-labelling. Our solution has ranked second in the two categories of the 2021 Pose Estimation Challenge organised by the European Space Agency and the Stanford University, achieving the lowest average error over the two categories., Comment: Accepted at ECCV 2022 AI4SPACE Workshop (https://aiforspace.github.io/2022/)

Spin-orbit interactions of light are ubiquitous in multiple branches of nanophotonics, including optical wave localization. In that framework, it is widely accepted that circularly polarized beams lead to spin-dependent apparent shifts of dipolar targets commonly referred to as optical mirages. In contrast, these optical mirages vanish when the illumination comes from a spinless beam such as a linearly polarized wave. Here we show that optical localization errors emerge for particles sustaining electric and magnetic dipolar response under the illumination of spinless beams. As an example, we calculate the optical mirage for the scattering by a high refractive index nanosphere under the illumination of a linearly polarized plane wave carrying null spin, orbital, and total angular momentum. Our results point to an overlooked interference between the electric and magnetic dipoles rather than the spin-orbit interactions of light as the origin for the tilted position of the nanosphere.

The possibility of creating and manipulating nanostructured materials encouraged the exploration of new strategies to control electromagnetic properties. Among the most intriguing nanostructures are those that respond differently to helical polarization, i.e., exhibit chirality. Here, we present a simple structure based on crossed elongated bars where light-handedness defines the dominating cross-section absorption or scattering, with a 200% difference from its counterpart (scattering or absorption). The proposed chiral system opens the way to enhanced coherent phonon excitation and detection. We theoretically propose a simple coherent phonon generation (time-resolved Brillouin scattering) experiment using circularly polarized light. In the reported structures, the generation of acoustic phonons is optimized by maximizing the absorption, while the detection is enhanced at the same wavelength -- and different helicity -- by engineering the scattering properties. The presented results constitute one of the first steps towards harvesting chirality effects in the design and optimization of efficient and versatile acoustoplasmonic transducers., Comment: 9 pages, 6 figures + 2 pages for Supplementary Info. with two figures

Light-harvesting complexes in natural photosynthetic systems, such as those in purple bacteria, consist of photo-reactive chromophores embedded in densely packed "antenna" systems organized in well-defined nanostructures. In the case of purple bacteria, the chromophore antennas are composed of natural J-aggregates such as bacteriochlorophylls and carotenoids. Inspired by the molecular composition of such biological systems, we create a library of organic materials composed of densely packed J-aggregates in a polymeric matrix, in which the matrix mimics a protein scaffold. This library of organic materials shows polaritonic properties which can be tuned from the visible to the infrared by choice of the model molecule. Inspired by the molecular architecture of the light-harvesting complexes of \textit{Rhodospirillum molischianum} bacteria, we study the light-matter interactions of J-aggregate-based nanorings with similar dimensions to the analogous natural nanoscale architectures. Electromagnetic simulations show that these nanorings of J-aggregates can act as resonators, with subwavelength confinement of light while concentrating the electric field in specific regions. These results open the door to bio-inspired building blocks for all-organic metamaterials while offering a new perspective on light-matter interactions at the nanoscale in densely packed organic matter in biological organisms including photosynthetic organelles., Comment: Including supplementary information

Optical Tamm States (OTS) are confined optical modes that can occur at the interface between two highly reflective structures. However, due to the strong reflectance required, their implemen-tation with highly processable and metal-free flexible materials has proven challenging. Herein, we develop the first structure supporting OTS based only on organic polymeric materials, demon-strating a photonic platform based on non-critical, widely available, and easily processable mate-rials. The structures fabricated present large areas and consist of a narrowband multi-layered polymeric Distributed Bragg Reflector (DBR) followed by a thin film of J-aggregate molecular exci-tonic material that can act as a highly reflective surface within a narrowband range. We take ad-vantage of the narrowband spectral response of the DBR and of the reflective molecular layer to tune the OTS band by varying the periodicity of the multilayer, opening the door for the fabrica-tion of OTS structures based on lightweight integrable excitonic devices with cost-effective proce-dures., Comment: 15 pages, 8 (3+5) figures, supplementary information included