Search

The measurement-based architecture is a paradigm of quantum computing, relying on the entanglement of a cluster of qubits and the measurements of a subset of it, conditioning the state of the unmeasured output qubits. While methods to map the gate model circuits into the measurement-based are already available via intermediate steps, we introduce a new paradigm for quantum compiling directly converting any quantum circuit to a class of graph states, independently from its size. Such method relies on the stabilizer formalism to describe the register of the input qubits. An equivalence class between graph states able to implement the same circuit is defined, giving rise to a gauge freedom when compiling in the MBQC frame. The graph state can be rebuilt from the circuit and the input by employing a set of graphical rules similar to the Feynman's ones. A system of equations describes the overall process. Compared to Measurement Calculus, the ancillary qubits are reduced by 50% on QFT and 75% on QAOA., Comment: 28 pages, 6 figures, 2 tables, 1 pseudo-algorithm

We introduce the LUNA neutron detector array developed for the investigation of the 13C(a,n)16O reaction towards its astrophysical s-process Gamow peak in the low-background environment of the Laboratori Nazionali del Gran Sasso (LNGS). Eighteen 3He counters are arranged in two different configurations (in a vertical and a horizontal orientation) to optimize neutron detection effciency, target handling and target cooling over the investigated energy range Ea;lab = 300 - 400 keV (En = 2.2 - 2.6 MeV in emitted neutron energy). As a result of the deep underground location, the passive shielding of the setup and active background suppression using pulse shape discrimination, we reached a total background rate of 1.23 +- 0.12 counts/hour. This resulted in an improvement of two orders of magnitude over the state of the art allowing a direct measurement of the 13C(a,n)16O cross-section down to Ea;lab = 300 keV. The absolute neutron detection efficiency of the setup was determined using the 51V(p,n)51Cr reaction and an AmBe radioactive source, and completed with a Geant4 simulation. We determined a (34+-3) % and (38+-3) % detection efficiency for the vertical and horizontal configurations, respectively, for En = 2.4 MeV neutrons., Comment: 8 pages, 5 figures, to be published in NIMA

When combined with In-Context Learning, a technique that enables models to adapt to new tasks by incorporating task-specific examples or demonstrations directly within the input prompt, autoregressive language models have achieved good performance in a wide range of tasks and applications. However, this combination has not been properly explored in the context of named entity recognition, where the structure of this task poses unique challenges. We propose RENER (Retrieval-Enhanced Named Entity Recognition), a technique for named entity recognition using autoregressive language models based on In-Context Learning and information retrieval techniques. When presented with an input text, RENER fetches similar examples from a dataset of training examples that are used to enhance a language model to recognize named entities from this input text. RENER is modular and independent of the underlying language model and information retrieval algorithms. Experimental results show that in the CrossNER collection we achieve state-of-the-art performance with the proposed technique and that information retrieval can increase the F-score by up to 11 percentage points., Comment: 13 pages, 6 figures, 3 tables

Recent advancements in generative models have revolutionized the field of artificial intelligence, enabling the creation of highly-realistic and detailed images. In this study, we propose a novel Mask Conditional Text-to-Image Generative Model (MCGM) that leverages the power of conditional diffusion models to generate pictures with specific poses. Our model builds upon the success of the Break-a-scene [1] model in generating new scenes using a single image with multiple subjects and incorporates a mask embedding injection that allows the conditioning of the generation process. By introducing this additional level of control, MCGM offers a flexible and intuitive approach for generating specific poses for one or more subjects learned from a single image, empowering users to influence the output based on their requirements. Through extensive experimentation and evaluation, we demonstrate the effectiveness of our proposed model in generating high-quality images that meet predefined mask conditions and improving the current Break-a-scene generative model., Comment: 17 pages, 13 figures, presented at the 5th International Conference on Artificial Intelligence and Machine Learning (CAIML 2024)

The goal of style transfer is, given a content image and a style source, generating a new image preserving the content but with the artistic representation of the style source. Most of the state-of-the-art architectures use transformers or diffusion-based models to perform this task, despite the heavy computational burden that they require. In particular, transformers use self- and cross-attention layers which have large memory footprint, while diffusion models require high inference time. To overcome the above, this paper explores a novel design of Mamba, an emergent State-Space Model (SSM), called Mamba-ST, to perform style transfer. To do so, we adapt Mamba linear equation to simulate the behavior of cross-attention layers, which are able to combine two separate embeddings into a single output, but drastically reducing memory usage and time complexity. We modified the Mamba's inner equations so to accept inputs from, and combine, two separate data streams. To the best of our knowledge, this is the first attempt to adapt the equations of SSMs to a vision task like style transfer without requiring any other module like cross-attention or custom normalization layers. An extensive set of experiments demonstrates the superiority and efficiency of our method in performing style transfer compared to transformers and diffusion models. Results show improved quality in terms of both ArtFID and FID metrics. Code is available at https://github.com/FilippoBotti/MambaST.

Complex quantum circuits are constituted by combinations of quantum subroutines. The computation is possible as long as the quantum data encoding is consistent throughout the circuit. Despite its fundamental importance, the formalization of quantum data encoding has never been addressed systematically so far. We formalize the concept of quantum data encoding, namely the format providing a representation of a data set through a quantum state, as a distinct abstract layer with respect to the associated data loading circuit. We survey existing encoding methods and their respective strategies for classical-to-quantum exact and approximate data loading, for the quantum-to-classical extraction of information from states, and for quantum-to-quantum encoding conversion. Next, we show how major quantum algorithms find a natural interpretation in terms of data loading. For instance, the Quantum Fourier Transform is described as a quantum encoding converter, while the Quantum Amplitude Estimation as an extraction routine. The new conceptual framework is exemplified by considering its application to quantum-based Monte Carlo simulations, thus showcasing the power of the proposed formalism for the description of complex quantum circuits. Indeed, the approach clarifies the structure of complex quantum circuits and enables their efficient design.

We identify a noise model that ensures the functioning of an echo state network employing a gate-based quantum computer for reservoir computing applications. Energy dissipation induced by amplitude damping drastically improves the short-term memory capacity and expressivity of the network, by simultaneously providing fading memory and richer dynamics. There is an ideal dissipation rate that ensures the best operation of the echo state network around $\gamma\sim$ 0.03. Nevertheless, these beneficial effects are stable as the intensity of the applied noise increases. The improvement of the learning is confirmed by emulating a realistic noise model applied to superconducting qubits, paving the way for the application of reservoir computing methods in current non-fault-tolerant quantum computers., Comment: 10 pages, 7 figures

The advent of quantum computers has justified the development of quantum machine learning algorithms , based on the adaptation of the principles of machine learning to the formalism of qubits. Among such quantum algorithms, anomaly detection represents an important problem crossing several disciplines from cybersecurity, to fraud detection to particle physics. We summarize the key concepts involved in quantum computing, introducing the formal concept of quantum speed up. The survey provides a structured map of anomaly detection based on quantum machine learning. We have grouped existing algorithms according to the different learning methods, namely quantum supervised, quantum unsupervised and quantum reinforcement learning, respectively. We provide an estimate of the hardware resources to provide sufficient computational power in the future. The survey provides a systematic and compact understanding of the techniques belonging to each category. We eventually provide a discussion on the computational complexity of the learning methods in real application domains.

Coupling is an essential mechanism that drives complexity in natural systems, transforming single, non-interacting elements into intricate networks with rich physical properties. Here, we demonstrate a chip-scale coupled laser system that exhibits complex optical states impossible to achieve in an uncoupled system. We show that a pair of coupled semiconductor ring lasers spontaneously forms a frequency comb consisting of the hybridized modes of its coupled cavity, exhibiting a large number of phase-locked tones that anticross with one another. Experimental coherent waveform reconstruction reveals that the hybridized frequency comb manifests itself as pairs of bright and dark picosecond-long solitons circulating simultaneously. The dark and bright solitons exit the coupled cavity at the same time, leading to breathing bright solitons temporally overlapped with their dark soliton counterparts - a state inaccessible for a single, free-running laser. Our results demonstrate that the rules that govern allowable states of light can be broken by simply coupling elements together, paving the way for the design of more complex networks of coupled on-chip lasers., Comment: 13 pages, 3 figures

The CNO cycle is one of the most important nuclear energy sources in stars. At temperatures of hydrostatic H-burning (20 MK $<$ T $<$ 80 MK) the $^{17}$O(p,$\gamma$)$^{18}$F reaction rate is dominated by the poorly constrained 64.5~keV resonance. Here we report on the first direct measurements of its resonance strength and of the direct capture contribution at 142 keV, performed with a new high sensitivity setup at LUNA. The present resonance strength of $\omega\gamma_{(p, \gamma)}$\textsuperscript{bare} = (30 $\pm$ 6\textsubscript{stat} $\pm$ 2\textsubscript{syst})~peV is about a factor of 2 higher than the values in literature, leading to a $\Gamma$\textsubscript{p}\textsuperscript{bare} = (34 $\pm$ 7\textsubscript{stat} $\pm$ 3\textsubscript{syst})~neV, in agreement with LUNA result from the (p,$\alpha$) channel. Such agreement strengthen our understanding of the oxygen isotopic ratios measured in red giant stars and in O-rich presolar grains.

Quantum computers have proven to be effective in simulating many quantum systems. Simulating nuclear processes and state preparation poses significant challenges, even for traditional supercomputers. This study demonstrates the feasibility of a complete simulation of a nuclear transition, including the preparation of both ground and first excited states. To tackle the complexity of strong interactions between two and three nucleons, the states are modeled on the tritium nucleus. Both the initial and final states are represented using quantum circuits with variational quantum algorithms and inductive biases. Describing the spin-isospin states requires four qubits, and a parameterized quantum circuit that exploits a total of 16 parameters is initialized. The estimated energy has a relative error of approximately 2% for the ground state and about 10% for the first excited state of the system. The quantum computer simulation estimates the transition probability between the two states as a function of the dipole polarization angle. This work marks a first step towards leveraging digital quantum computers to simulate nuclear physics., Comment: 22 pages, 7 figures

Quantum many-body systems are of great interest for many research areas, including physics, biology and chemistry. However, their simulation is extremely challenging, due to the exponential growth of the Hilbert space with the system size, making it exceedingly difficult to parameterize the wave functions of large systems by using exact methods. Neural networks and machine learning in general are a way to face this challenge. For instance, methods like Tensor networks and Neural Quantum States are being investigated as promising tools to obtain the wave function of a quantum mechanical system. In this tutorial, we focus on a particularly promising class of deep learning algorithms. We explain how to construct a Physics-Informed Neural Network (PINN) able to solve the Schr\"odinger equation for a given potential, by finding its eigenvalues and eigenfunctions. This technique is unsupervised, and utilizes a novel computational method in a manner that is barely explored. PINNs are a deep learning method that exploits Automatic Differentiation to solve Integro-Differential Equations in a mesh-free way. We show how to find both the ground and the excited states. The method discovers the states progressively by starting from the ground state. We explain how to introduce inductive biases in the loss to exploit further knowledge of the physical system. Such additional constraints allow for a faster and more accurate convergence. This technique can then be enhanced by a smart choice of collocation points in order to take advantage of the mesh-free nature of the PINN. The methods are made explicit by applying them to the infinite potential well and the particle in a ring, a challenging problem to be learned by an Artificial Intelligence agent due to the presence of complex-valued eigenfunctions and degenerate states., Comment: 18 pages, 4 figures

Text-based person search (TBPS) is a problem that gained significant interest within the research community. The task is that of retrieving one or more images of a specific individual based on a textual description. The multi-modal nature of the task requires learning representations that bridge text and image data within a shared latent space. Existing TBPS systems face two major challenges. One is defined as inter-identity noise that is due to the inherent vagueness and imprecision of text descriptions and it indicates how descriptions of visual attributes can be generally associated to different people; the other is the intra-identity variations, which are all those nuisances e.g. pose, illumination, that can alter the visual appearance of the same textual attributes for a given subject. To address these issues, this paper presents a novel TBPS architecture named MARS (Mae-Attribute-Relation-Sensitive), which enhances current state-of-the-art models by introducing two key components: a Visual Reconstruction Loss and an Attribute Loss. The former employs a Masked AutoEncoder trained to reconstruct randomly masked image patches with the aid of the textual description. In doing so the model is encouraged to learn more expressive representations and textual-visual relations in the latent space. The Attribute Loss, instead, balances the contribution of different types of attributes, defined as adjective-noun chunks of text. This loss ensures that every attribute is taken into consideration in the person retrieval process. Extensive experiments on three commonly used datasets, namely CUHK-PEDES, ICFG-PEDES, and RSTPReid, report performance improvements, with significant gains in the mean Average Precision (mAP) metric w.r.t. the current state of the art.

The use of deep learning in physical sciences has recently boosted the ability of researchers to tackle physical systems where little or no analytical insight is available. Recently, the Physics-Informed Neural Networks (PINNs) have been introduced as one of the most promising tools to solve systems of differential equations guided by some physically grounded constraints. In the quantum realm, such approach paves the way to a novel approach to solve the Schroedinger equation for non-integrable systems. By following an unsupervised learning approach, we apply the PINNs to the anharmonic oscillator in which an interaction term proportional to the fourth power of the position coordinate is present. We compute the eigenenergies and the corresponding eigenfunctions while varying the weight of the quartic interaction. We bridge our solutions to the regime where both the perturbative and the strong coupling theory work, including the pure quartic oscillator. We investigate systems with real and imaginary frequency, laying the foundation for novel numerical methods to tackle problems emerging in quantum field theory., Comment: 18 pages, 10 figures. Comments are welcome

Due to the limitations of current optical and sensor technologies and the high cost of updating them, the spectral and spatial resolution of satellites may not always meet desired requirements. For these reasons, Remote-Sensing Single-Image Super-Resolution (RS-SISR) techniques have gained significant interest. In this paper, we propose Swin2-MoSE model, an enhanced version of Swin2SR. Our model introduces MoE-SM, an enhanced Mixture-of-Experts (MoE) to replace the Feed-Forward inside all Transformer block. MoE-SM is designed with Smart-Merger, and new layer for merging the output of individual experts, and with a new way to split the work between experts, defining a new per-example strategy instead of the commonly used per-token one. Furthermore, we analyze how positional encodings interact with each other, demonstrating that per-channel bias and per-head bias can positively cooperate. Finally, we propose to use a combination of Normalized-Cross-Correlation (NCC) and Structural Similarity Index Measure (SSIM) losses, to avoid typical MSE loss limitations. Experimental results demonstrate that Swin2-MoSE outperforms SOTA by up to 0.377 ~ 0.958 dB (PSNR) on task of 2x, 3x and 4x resolution-upscaling (Sen2Venus and OLI2MSI datasets). We show the efficacy of Swin2-MoSE, applying it to a semantic segmentation task (SeasoNet dataset). Code and pretrained are available on https://github.com/IMPLabUniPr/swin2-mose/tree/official_code

Current state-of-the-art two-stage models on instance segmentation task suffer from several types of imbalances. In this paper, we address the Intersection over the Union (IoU) distribution imbalance of positive input Regions of Interest (RoIs) during the training of the second stage. Our Self-Balanced R-CNN (SBR-CNN), an evolved version of the Hybrid Task Cascade (HTC) model, brings brand new loop mechanisms of bounding box and mask refinements. With an improved Generic RoI Extraction (GRoIE), we also address the feature-level imbalance at the Feature Pyramid Network (FPN) level, originated by a non-uniform integration between low- and high-level features from the backbone layers. In addition, the redesign of the architecture heads toward a fully convolutional approach with FCC further reduces the number of parameters and obtains more clues to the connection between the task to solve and the layers used. Moreover, our SBR-CNN model shows the same or even better improvements if adopted in conjunction with other state-of-the-art models. In fact, with a lightweight ResNet-50 as backbone, evaluated on COCO minival 2017 dataset, our model reaches 45.3% and 41.5% AP for object detection and instance segmentation, with 12 epochs and without extra tricks. The code is available at https://github.com/IMPLabUniPr/mmdetection/tree/sbr_cnn

Nowadays, deep learning models have reached incredible performance in the task of image generation. Plenty of literature works address the task of face generation and editing, with human and automatic systems that struggle to distinguish what's real from generated. Whereas most systems reached excellent visual generation quality, they still face difficulties in preserving the identity of the starting input subject. Among all the explored techniques, Semantic Image Synthesis (SIS) methods, whose goal is to generate an image conditioned on a semantic segmentation mask, are the most promising, even though preserving the perceived identity of the input subject is not their main concern. Therefore, in this paper, we investigate the problem of identity preservation in face image generation and present an SIS architecture that exploits a cross-attention mechanism to merge identity, style, and semantic features to generate faces whose identities are as similar as possible to the input ones. Experimental results reveal that the proposed method is not only suitable for preserving the identity but is also effective in the face recognition adversarial attack, i.e. hiding a second identity in the generated faces., Comment: Paper accepted at CVPR 2024 Biometrics Workshop

Semantic Image Synthesis (SIS) is among the most popular and effective techniques in the field of face generation and editing, thanks to its good generation quality and the versatility is brings along. Recent works attempted to go beyond the standard GAN-based framework, and started to explore Diffusion Models (DMs) for this task as these stand out with respect to GANs in terms of both quality and diversity. On the other hand, DMs lack in fine-grained controllability and reproducibility. To address that, in this paper we propose a SIS framework based on a novel Latent Diffusion Model architecture for human face generation and editing that is both able to reproduce and manipulate a real reference image and generate diversity-driven results. The proposed system utilizes both SPADE normalization and cross-attention layers to merge shape and style information and, by doing so, allows for a precise control over each of the semantic parts of the human face. This was not possible with previous methods in the state of the art. Finally, we performed an extensive set of experiments to prove that our model surpasses current state of the art, both qualitatively and quantitatively.

A bimetallic Au-Pd catalyst supported on ceriazirconia with Au:Pd molar ratio 0.8 has been synthesized using a simultaneous deposition-precipitation method and oxidized at 250, 450, and 700 $^\circ$C in order to modify its particle size, nanostructure, and composition. Combined Xray energy dispersive spectroscopy and Xray photoelectron spectroscopy analysis clearly evidence that the bimetallic Au-Pd catalyst oxidized at 250 $^\circ$C is made up of a mixture of monometallic Au and Pd and bimetallic Au-Pd nanoparticles with Au:Pd ratios varying over a wide range. Increasing oxidation temperature leads to a stronger interaction between Au and Pd. Meanwhile, a slight increase of particle size and a narrowing of the Au:Pd ratio in the bimetallic nanoparticles take place. Compared with titania and activated carbon supports, the resistance against sintering at high temperatures of Au-Pd metal particles supported on ceriazirconia is proven to be higher. A synergistic effect has been observed for selective oxidation of benzyl alcohol on these catalysts. The catalytic activity decreases only slightly after oxidation at 450 $^\circ$C. However, oxidation at 700 $^\circ$C results in much lower catalytic activity. Migration of Pd onto Au particles during oxidation of benzyl alcohol enhances the catalytic activity of a physical mixture of monometallic Au and Pd supported on ceriazirconia catalysts. This fact, jointly with an analysis of the intrinsic activity, reveals the influence of the actual nature of Au-Pd interactions in the bimetallic particles, which points to higher activity of Au@Pd or Au@Pd@Pd nanostructures on ceria-zirconia support., Comment: 12 pages, 8 figures

Despite the ongoing progress in integrated optical frequency comb technology, compact sources of short bright pulses in the mid-infrared wavelength range from 3 {\mu}m to 12 {\mu}m so far remained beyond reach. The state-of-the-art ultrafast pulse emitters in the mid-infrared are complex, bulky, and inefficient systems based on the downconversion of near-infrared or visible pulsed laser sources. Here we show a purely DC-driven semiconductor laser chip that generates one picosecond solitons at the center wavelength of 8.3 {\mu}m at GHz repetition rates. The soliton generation scheme is akin to that of passive nonlinear Kerr resonators. It relies on a fast bistability in active nonlinear laser resonators, unlike traditional passive mode-locking which relies on saturable absorbers or active mode-locking by gain modulation in semiconductor lasers. Monolithic integration of all components - drive laser, active ring resonator, coupler, and pump filter - enables turnkey generation of bright solitons that remain robust for hours of continuous operation without active stabilization. Such devices can be readily produced at industrial laser foundries using standard fabrication protocols. Our work unifies the physics of active and passive microresonator frequency combs, while simultaneously establishing a technology for nonlinear integrated photonics in the mid-infrared.

Reservoir computing is a versatile paradigm in computational neuroscience and machine learning, that exploits the non-linear dynamics of a dynamical system - the reservoir - to efficiently process time-dependent information. Since its introduction, it has exhibited remarkable capabilities in various applications. As widely known, classes of reservoir computers serve as universal approximators of functionals with fading memory. The construction of such universal classes often appears context-specific, but in fact, they follow the same principles. Here we present a unified theoretical framework and we propose a ready-made setting to secure universality. We test the result in the arising context of quantum reservoir computing.The analysis sheds light on a unified view of classical and quantum reservoir computing., Comment: 13 pages, 1 figure

Coherent multistatic radio imaging represents a pivotal opportunity for forthcoming wireless networks, which involves distributed nodes cooperating to achieve accurate sensing resolution and robustness. This paper delves into cooperative coherent imaging for vehicular radar networks. Herein, multiple radar-equipped vehicles cooperate to improve collective sensing capabilities and address the fundamental issue of distinguishing weak targets in close proximity to strong ones, a critical challenge for vulnerable road users protection. We prove the significant benefits of cooperative coherent imaging in the considered automotive scenario in terms of both probability of correct detection, evaluated considering several system parameters, as well as resolution capabilities, showcased by a dedicated experimental campaign wherein the collaboration between two vehicles enables the detection of the legs of a pedestrian close to a parked car. Moreover, as \textit{coherent} processing of several sensors' data requires very tight accuracy on clock synchronization and sensor's positioning -- referred to as \textit{phase synchronization} -- (such that to predict sensor-target distances up to a fraction of the carrier wavelength), we present a general three-step cooperative multistatic phase synchronization procedure, detailing the required information exchange among vehicles in the specific automotive radar context and assessing its feasibility and performance by hybrid Cram\'er-Rao bound., Comment: 13 pages

The $\mathrm{^{20}Ne(p, \gamma)^{21}Na}$ reaction is the slowest in the NeNa cycle and directly affects the abundances of the Ne and Na isotopes in a variety of astrophysical sites. Here we report the measurement of its direct capture contribution, for the first time below $E\rm_{cm} = 352$~keV, and of the contribution from the $E^{\rm }_{cm} = 368$~keV resonance, which dominates the reaction rate at $T=0.03-1.00$~GK. The experiment was performed deep underground at the Laboratory for Underground Nuclear Astrophysics, using a high-intensity proton beam and a windowless neon gas target. Prompt $\gamma$ rays from the reaction were detected with two high-purity germanium detectors. We obtain a resonance strength $\omega \gamma~=~(0.112 \pm 0.002_{\rm stat}~\pm~0.005_{\rm sys})$~meV, with an uncertainty a factor of $3$ smaller than previous values. Our revised reaction rate is 20\% lower than previously adopted at $T < 0.1$~GK and agrees with previous estimates at temperatures $T \geq 0.1$~GK. Initial astrophysical implications are presented., Comment: 8 pages, 4 figures, accepted to PRC (letter)

The ${}^{12}\mathrm{C}/{}^{13}\mathrm{C}$ ratio is a significant indicator of nucleosynthesis and mixing processes during hydrogen burning in stars. Its value mainly depends on the relative rates of the ${}^{12}\mathrm{C}(p,\gamma){}^{13}\mathrm{N}$ and ${}^{13}\mathrm{C}(p,\gamma){}^{14}\mathrm{N}$ reactions. Both reactions have been studied at the Laboratory for Underground Nuclear Astrophysics (LUNA) in Italy down to the lowest energies to date ($E_\mathrm{c.m.} = 60\,\mathrm{keV}$) reaching for the first time the high energy tail of hydrogen burning in the shell of giant stars. Our cross sections, obtained with both prompt $\gamma$-ray detection and activation measurements, are the most precise to date with overall systematic uncertainties of $7-8\%$. Compared with most of the literature, our results are systematically lower, by $25\%$ for the ${}^{12}\mathrm{C}(p,\gamma){}^{13}\mathrm{N}$ reaction and by $30\%$ for ${}^{13}\mathrm{C}(p,\gamma){}^{14}\mathrm{N}$. We provide the most precise value up to now of $(3.6 \pm 0.4)$ in the $20-140\,\mathrm{MK}$ range for the lowest possible ${}^{12}\mathrm{C}/{}^{13}\mathrm{C}$ ratio that can be produced during H burning in giant stars., Comment: Submitted to Phys. Rev. Lett

In semantic image synthesis the state of the art is dominated by methods that use customized variants of the SPatially-Adaptive DE-normalization (SPADE) layers, which allow for good visual generation quality and editing versatility. By design, such layers learn pixel-wise modulation parameters to de-normalize the generator activations based on the semantic class each pixel belongs to. Thus, they tend to overlook global image statistics, ultimately leading to unconvincing local style editing and causing global inconsistencies such as color or illumination distribution shifts. Also, SPADE layers require the semantic segmentation mask for mapping styles in the generator, preventing shape manipulations without manual intervention. In response, we designed a novel architecture where cross-attention layers are used in place of SPADE for learning shape-style correlations and so conditioning the image generation process. Our model inherits the versatility of SPADE, at the same time obtaining state-of-the-art generation quality, as well as improved global and local style transfer. Code and models available at https://github.com/TFonta/CA2SIS., Comment: Code and models available at https://github.com/TFonta/CA2SIS The paper is under consideration at Computer Vision and Image Understanding

Semantic image synthesis (SIS) refers to the problem of generating realistic imagery given a semantic segmentation mask that defines the spatial layout of object classes. Most of the approaches in the literature, other than the quality of the generated images, put effort in finding solutions to increase the generation diversity in terms of style i.e. texture. However, they all neglect a different feature, which is the possibility of manipulating the layout provided by the mask. Currently, the only way to do so is manually by means of graphical users interfaces. In this paper, we describe a network architecture to address the problem of automatically manipulating or generating the shape of object classes in semantic segmentation masks, with specific focus on human faces. Our proposed model allows embedding the mask class-wise into a latent space where each class embedding can be independently edited. Then, a bi-directional LSTM block and a convolutional decoder output a new, locally manipulated mask. We report quantitative and qualitative results on the CelebMask-HQ dataset, which show our model can both faithfully reconstruct and modify a segmentation mask at the class level. Also, we show our model can be put before a SIS generator, opening the way to a fully automatic generation control of both shape and texture. Code available at https://github.com/TFonta/Semantic-VAE., Comment: Preprint, accepted for publication at ICIAP 2023

Networked sensing refers to the capability of properly orchestrating multiple sensing terminals to enhance specific figures of merit, e.g., positioning accuracy or imaging resolution. Regarding radio-based sensing, it is essential to understand \textit{when} and \textit{how} sensing terminals should be orchestrated, namely the best cooperation that trades between performance and cost (e.g., energy consumption, communication overhead, and complexity). This paper addresses networked sensing from a physics-driven perspective, aiming to provide a general theoretical benchmark to evaluate its \textit{imaging} performance bounds and to guide the sensing orchestration accordingly. Diffraction tomography theory (DTT) is the method to quantify the imaging resolution of any radio sensing experiment from inspection of its spectral (or wavenumber) content. In networked sensing, the image formation is based on the back-projection integral, valid for any network topology and physical configuration of the terminals. The \textit{wavefield networked sensing} is a framework in which multiple sensing terminals are orchestrated during the acquisition process to maximize the imaging quality (resolution and grating lobes suppression) by pursuing the deceptively simple \textit{wavenumber tessellation principle}. We discuss all the cooperation possibilities between sensing terminals and possible killer applications. Remarkably, we show that the proposed method allows obtaining high-quality images of the environment in limited bandwidth conditions, leveraging the coherent combination of multiple multi-static low-resolution images., Comment: 13 pages

Recent years witnessed rapid progress of chip-scale integrated optical frequency comb sources. Among them, two classes are particularly significant -- semiconductor Fabry-Per\'{o}t lasers and passive ring Kerr microresonators. Here, we merge the two technologies in a ring semiconductor laser and demonstrate a new paradigm for free-running soliton formation, called Nozaki-Bekki soliton. These dissipative waveforms emerge in a family of traveling localized dark pulses, known within the famed complex Ginzburg-Landau equation. We show that Nozaki-Bekki solitons are structurally-stable in a ring laser and form spontaneously with tuning of the laser bias -- eliminating the need for an external optical pump. By combining conclusive experimental findings and a complementary elaborate theoretical model, we reveal the salient characteristics of these solitons and provide a guideline for their generation. Beyond the fundamental soliton circulating inside the ring laser, we demonstrate multisoliton states as well, verifying their localized nature and offering an insight into formation of soliton crystals. Our results consolidate a monolithic electrically-driven platform for direct soliton generation and open a door for a new research field at the junction of laser multimode dynamics and Kerr parametric processes.

The ability to understand the surrounding scene is of paramount importance for Autonomous Vehicles (AVs). This paper presents a system capable to work in an online fashion, giving an immediate response to the arise of anomalies surrounding the AV, exploiting only the videos captured by a dash-mounted camera. Our architecture, called MOVAD, relies on two main modules: a Short-Term Memory Module to extract information related to the ongoing action, implemented by a Video Swin Transformer (VST), and a Long-Term Memory Module injected inside the classifier that considers also remote past information and action context thanks to the use of a Long-Short Term Memory (LSTM) network. The strengths of MOVAD are not only linked to its excellent performance, but also to its straightforward and modular architecture, trained in a end-to-end fashion with only RGB frames with as less assumptions as possible, which makes it easy to implement and play with. We evaluated the performance of our method on Detection of Traffic Anomaly (DoTA) dataset, a challenging collection of dash-mounted camera videos of accidents. After an extensive ablation study, MOVAD is able to reach an AUC score of 82.17\%, surpassing the current state-of-the-art by +2.87 AUC. Our code will be available on https://github.com/IMPLabUniPr/movad/tree/movad_vad

Graph states are the main computational building blocks of measurement-based computation and a useful tool for error correction in the gate model architecture. The graph states form a class of quantum states which are eigenvectors for the abelian group of stabilizer operators. They own topological properties, arising from their graph structure, including the presence of highly connected nodes, called hubs. Starting from hub nodes, we demonstrate how to efficiently express a graph state through the generators of the stabilizer group. We provide examples by expressing the ring and the star topology, for which the number of stabilizers reduces from n to n/2, and from n to 1, respectively. We demonstrate that the graph states can be generated by a subgroup of the stabilizer group. Therefore, we provide an algebraic framework to manipulate the graph states with a reduced number of stabilizers., Comment: 5 pages (excluding references), 2 figures (each with 2 sub-figures)

Quantum reservoir computing is a class of quantum machine learning algorithms involving a reservoir of an echo state network based on a register of qubits, but the dependence of its memory capacity on the hyperparameters is still rather unclear. In order to maximize its accuracy in time--series predictive tasks, we investigate the relation between the memory of the network and the reset rate of the evolution of the quantum reservoir. We benchmark the network performance by three non--linear maps with fading memory on IBM quantum hardware. The memory capacity of the quantum reservoir is maximized for central values of the memory reset rate in the interval [0,1]. As expected, the memory capacity increases approximately linearly with the number of qubits. After optimization of the memory reset rate, the mean squared errors of the predicted outputs in the tasks may decrease by a factor ~1/5 with respect to previous implementations., Comment: 21 pages, 5 figures