Search

In unsupervised anomaly detection (UAD) research, while state-of-the-art models have reached a saturation point with extensive studies on public benchmark datasets, they adopt large-scale tailor-made neural networks (NN) for detection performance or pursued unified models for various tasks. Towards edge computing, it is necessary to develop a computationally efficient and scalable solution that avoids large-scale complex NNs. Motivated by this, we aim to optimize the UAD performance with minimal changes to NN settings. Thus, we revisit the reconstruction-by-inpainting approach and rethink to improve it by analyzing strengths and weaknesses. The strength of the SOTA methods is a single deterministic masking approach that addresses the challenges of random multiple masking that is inference latency and output inconsistency. Nevertheless, the issue of failure to provide a mask to completely cover anomalous regions is a remaining weakness. To mitigate this issue, we propose Feature Attenuation of Defective Representation (FADeR) that only employs two MLP layers which attenuates feature information of anomaly reconstruction during decoding. By leveraging FADeR, features of unseen anomaly patterns are reconstructed into seen normal patterns, reducing false alarms. Experimental results demonstrate that FADeR achieves enhanced performance compared to similar-scale NNs. Furthermore, our approach exhibits scalability in performance enhancement when integrated with other single deterministic masking methods in a plug-and-play manner., Comment: 11 pages, 6 figures, 5 tables

The proliferation of edge devices necessitates efficient computational architectures for lightweight tasks, particularly deep neural network (DNN) inference. Traditional NPUs, though effective for such operations, face challenges in power, cost, and area when integrated into lightweight edge devices. The RISC-V architecture, known for its modularity and open-source nature, offers a viable alternative. This paper introduces the RISC-V R-extension, a novel approach to enhancing DNN process efficiency on edge devices. The extension features rented-pipeline stages and architectural pipeline registers (APR), which optimize critical operation execution, thereby reducing latency and memory access frequency. Furthermore, this extension includes new custom instructions to support these architectural improvements. Through comprehensive analysis, this study demonstrates the boost of R-extension in edge device processing, setting the stage for more responsive and intelligent edge applications., Comment: 6 pages, 6 figures, ICAIIC 2024

We present a family of charged rotating wormhole solutions to the Einstein-Maxwell equations, supported by anisotropic matter fields. We first revisit the charged static cases and analyze the conditions for the solution to represent a wormhole geometry. The rotating geometry is obtained by applying the Newman-Janis algorithm to the static geometry. We show the solutions to Maxwell equations in detail. We believe that our wormhole geometry offers a geometric realization corresponding to the concept of 'charge without charge'., Comment: 24 pages, 3 figures

Materials potentially hosting Kitaev spin-liquid states are considered crucial for realizing topological quantum computing. However, the intricate nature of spin interactions within these materials complicates the precise measurement of low-energy spin excitations indicative of fractionalized excitations. Using Na$_{2}$Co$_2$TeO$_{6}$ as an example, we study these low-energy spin excitations using the time-resolved resonant elastic x-ray scattering (tr-REXS). Our observations unveil remarkably slow spin dynamics at the magnetic peak, whose recovery timescale is several nanoseconds. This timescale aligns with the extrapolated spin gap of $\sim$ 1 $\mu$eV, obtained by density matrix renormalization group (DMRG) simulations in the thermodynamic limit. The consistency demonstrates the efficacy of tr-REXS in discerning low-energy spin gaps inaccessible to conventional spectroscopic techniques., Comment: 9 pages, 6 figures

We introduce ECLAIR (Extended Classification of Lidar for AI Recognition), a new outdoor large-scale aerial LiDAR dataset designed specifically for advancing research in point cloud semantic segmentation. As the most extensive and diverse collection of its kind to date, the dataset covers a total area of 10$km^2$ with close to 600 million points and features eleven distinct object categories. To guarantee the dataset's quality and utility, we have thoroughly curated the point labels through an internal team of experts, ensuring accuracy and consistency in semantic labeling. The dataset is engineered to move forward the fields of 3D urban modeling, scene understanding, and utility infrastructure management by presenting new challenges and potential applications. As a benchmark, we report qualitative and quantitative analysis of a voxel-based point cloud segmentation approach based on the Minkowski Engine., Comment: 11 pages, 7 figures

The piezoelectricity of group IV monochalcogenides (MXs, with M = Ge, Sn and X = S, Se) has attracted much attention due to their substantially higher piezoelectric coefficients compared to other 2D materials. However, with increasing layer number, their piezoelectricity rapidly disappears due to the antiferroelectric stacking order, severely limiting their practical applications. Using first-principles calculations, we investigated the piezoelectricity of MXs with the ferroelectric AA stacking configuration, which has recently been stabilized in experiments. We found that AA-stacked MXs have a ferroelectric ground state with the smallest lattice constant among other stacking configurations, resulting in a giant piezoelectric coefficient, which is the first demonstration of a strategy where the piezoelectric coefficients can increase with the number of layers. This can be attributed to a strong negative correlation between the lattice constant along the armchair direction and the piezoelectric coefficient, and spontaneous compressive strain stabilized in ferroelectric AA stacking configuration., Comment: 16 pages and 5 figures for main manuscript. 9pages and 3 figures for SI

We study the collision dynamics of a spinning cue ball approaching a static object ball with equal mass on a plane, common in billiards. While typical collisions in billiards are nearly perfectly elastic, with a restitution coefficient close to 1 and low friction, we explore three deviations from ideal elastic collisions: The non-elastic nature, the friction effects between the balls during collision, the friction between the ball and the table. We describe the detailed collision outcomes, emphasizing the importance of considering frictions. We reveal that friction, both between the balls and with the table, significantly influences the post-collision motions, deviating from the expectations of a purely elastic collision. The insights gained contribute to a better understanding of ball dynamics, impacting strategies and gameplay in billiards., Comment: 19 pages, 3 figures

We study the issue of temperature in a steady system around a black hole event horizon, contrasting it with the appearance of divergence in a thermal equilibrium system. We focus on a spherically symmetric system governed by general relativity, particularly examining the steady state with radial heat conduction. Employing an appropriate approximation, we derive exact solutions that illuminate the behaviors of number density, local temperature, and heat in the proximity of a black hole. We demonstrate that a carefully regulated heat inflow can maintain finite local temperatures at the black hole event horizon, even without considering the back-reaction of matter. This discovery challenges conventional expectations that the local temperature near the event horizon diverges in scenarios of thermal equilibrium. This implications shows that there's an intricate connection between heat and gravity in the realm of black hole thermodynamics., Comment: 18 pages, 1 figures

Employing small angle X-ray scattering (SAXS), we explore the conditions under which the assembly of gold nanoparticles (AuNPs) grafted with the thermo-sensitive polymer Poly(N-isopropylacrylamide) (PNIPAM) emerges. We find that short-range order assembly emerges by combining the addition of electrolytes or poly-electrolytes with raising the temperature of the suspensions above the lower-critical solution temperature (LCST) of PNIPAM. Our results show that the longer the PNIPAM chain is, the better organization in the assembled clusters. Interestingly, without added electrolytes, there is no evidence of AuNP assembly as a function of temperature, although untethered PNIPAM is known to undergo a coil-to-globule transition above its LCST. This study demonstrates another approach to assembling potential thermo-sensitive nanostructures for devices by leveraging the unique properties of PNIPAM., Comment: Published at ACS Macro Letters, DOI - https://doi.org/10.1021/acsmacrolett.3c00617

We study thermodynamics of a heat-conducting ideal gas system. The study is based on i) the first law of thermodynamics from action formulation which expects heat-dependence of energy density and ii) the existence condition of a (local) Lorentz boost between an Eckart observer and a Landau-Lifshitz observer--a condition that extends the stability criterion of thermal equilibrium. The implications of these conditions include: i) Heat contributes to the energy density through the combination $q/n\Theta^2$ where $q$, $n$, and $\Theta$ represent heat, the number density, and the temperature, respectively. ii) The energy density has a unique minimum at $q=0$. iii) The temperature upper bound suppresses the heat dependence of the energy density inverse quadratically. This result explains why the expected heat dependence is difficult to observe in ordinary situation thermodynamics., Comment: 9 pages, 2 figure, some changes, to appear in Heliyon

Due to scarcity of anomaly situations in the early manufacturing stage, an unsupervised anomaly detection (UAD) approach is widely adopted which only uses normal samples for training. This approach is based on the assumption that the trained UAD model will accurately reconstruct normal patterns but struggles with unseen anomalous patterns. To enhance the UAD performance, reconstruction-by-inpainting based methods have recently been investigated, especially on the masking strategy of suspected defective regions. However, there are still issues to overcome: 1) time-consuming inference due to multiple masking, 2) output inconsistency by random masking strategy, and 3) inaccurate reconstruction of normal patterns when the masked area is large. Motivated by this, we propose a novel reconstruction-by-inpainting method, dubbed Excision And Recovery (EAR), that features single deterministic masking based on the ImageNet pre-trained DINO-ViT and visual obfuscation for hint-providing. Experimental results on the MVTec AD dataset show that deterministic masking by pre-trained attention effectively cuts out suspected defective regions and resolve the aforementioned issues 1 and 2. Also, hint-providing by mosaicing proves to enhance the UAD performance than emptying those regions by binary masking, thereby overcomes issue 3. Our approach achieves a high UAD performance without any change of the neural network structure. Thus, we suggest that EAR be adopted in various manufacturing industries as a practically deployable solution., Comment: 10 pages, 5 figures, 5 tables

In quantum machine learning, algorithms with parameterized quantum circuits (PQC) based on a hardware-efficient ansatz (HEA) offer the potential for speed-ups over traditional classical algorithms. While much attention has been devoted to supervised learning tasks, unsupervised learning using PQC remains relatively unexplored. One promising approach within quantum machine learning involves optimizing fewer parameters in PQC than in its classical counterparts, under the assumption that a sub-optimal solution exists within the Hilbert space. In this paper, we introduce the Variational Quantum Approximate Spectral Clustering (VQASC) algorithm - a NISQ-compatible method that requires optimization of fewer parameters than the system size, N, traditionally required in classical problems. We present numerical results from both synthetic and real-world datasets. Furthermore, we propose a descriptor, complemented by numerical analysis, to identify an appropriate ansatz circuit tailored for VQASC., Comment: 21 pages, 6 figures

Unsupervised anomaly detection (UAD) is a widely adopted approach in industry due to rare anomaly occurrences and data imbalance. A desirable characteristic of an UAD model is contained generalization ability which excels in the reconstruction of seen normal patterns but struggles with unseen anomalies. Recent studies have pursued to contain the generalization capability of their UAD models in reconstruction from different perspectives, such as design of neural network (NN) structure and training strategy. In contrast, we note that containing of generalization ability in reconstruction can also be obtained simply from steep-shaped loss landscape. Motivated by this, we propose a loss landscape sharpening method by amplifying the reconstruction loss, dubbed Loss AMPlification (LAMP). LAMP deforms the loss landscape into a steep shape so the reconstruction error on unseen anomalies becomes greater. Accordingly, the anomaly detection performance is improved without any change of the NN architecture. Our findings suggest that LAMP can be easily applied to any reconstruction error metrics in UAD settings where the reconstruction model is trained with anomaly-free samples only., Comment: 5 pages, 4 figures, 2 tables

The challenge of optimal design of experiments (DOE) pervades materials science, physics, chemistry, and biology. Bayesian optimization has been used to address this challenge in vast sample spaces, although it requires framing experimental campaigns through the lens of maximizing some observable. This framing is insufficient for epistemic research goals that seek to comprehensively analyze a sample space, without an explicit scalar objective (e.g., the characterization of a wafer or sample library). In this work, we propose a flexible formulation of scientific value that recasts a dataset of input conditions and higher-dimensional observable data into a continuous, scalar metric. Intuitively, the scientific value function measures where observables change significantly, emulating the perspective of experts driving an experiment, and can be used in collaborative analysis tools or as an objective for optimization techniques. We demonstrate this technique by exploring simulated phase boundaries from different observables, autonomously driving a variable temperature measurement of a ferroelectric material, and providing feedback from a nanoparticle synthesis campaign. The method is seamlessly compatible with existing optimization tools, can be extended to multi-modal and multi-fidelity experiments, and can integrate existing models of an experimental system. Because of its flexibility, it can be deployed in a range of experimental settings for autonomous or accelerated experiments.

Hypothesis: Introducing charged terminal groups to polymers that graft nanoparticles enables Coulombic control over their assembly by tuning the pH and salinity of aqueous suspensions. Experiments: Gold nanoparticles (AuNPs) are grafted with poly(ethylene glycol) (PEG) terminated with CH3 (charge-neutral), COOH (negatively charged), or NH2 (positively charged) groups. The nanoparticles are characterized using dynamic light scattering, zeta-potential, and thermal gravimetric analysis. Liquid surface X-ray reflectivity (XR) and grazing incidence small-angle X-ray scattering (GISAXS) techniques are employed to determine the density profile and in-plane structure of the AuNP assembly across and on the aqueous surface. Findings: The assembly of PEG-AuNPs at the liquid/vapor interface can be tuned by adjusting pH or salinity, particularly for COOH terminals. However, the effect is less pronounced for NH2 terminals. These distinct assembly behaviors are attributed to the overall charge of PEG-AuNPs and the conformation of PEG. The COOH-PEG corona is the most compact, resulting in smaller superlattice constants. The net charge per particle depends not only on the PEG terminal groups but also on the cation sequestration of PEG and the intrinsic negative charge of the AuNP surface. NH2-PEG, due to its closeness to overall charge neutrality and the presence of hydrogen bonding, enables the assembly of NH2-PEG-AuNPs more readily., Comment: Published in the Journal of Colloid and Interface Science, DOI - 10.1016/j.jcis.2023.07.095

We investigate the steady state of heat conduction in general relativity using a variational approach for two-fluid dynamics. We adopt coordinates based on the Landau-Lifschitz observer because it allows us to describe thermodynamics with heat, formulated in the Eckart decomposition, on a static geometry. Through our analysis, we demonstrate that the stability condition of a thermal equilibrium state arises from the fundamental principle that heat cannot propagate faster than the speed of light. We then formulate the equations governing steady-state heat conduction and introduce a binormal equilibrium condition that the Tolman temperature gradient holds for the directions orthogonal to the heat flow. As an example, we consider radial heat conductions in a spherically symmetric spacetime. We find that the total diffusion over a spherical surface satisfies a red-shifted form, $J(r) \sqrt{-g_{tt}} =$ constant. We also discuss the behavior of local temperature around an event horizon and specify the condition that the local temperature is finite there., Comment: 16 pages, 1 figure, minor corrections, to appear in PTEP

In modeling relativistic thermodynamics, we frequently regard the particle number as a conserved quantity. The number conservation law, which comes from the requirement that the pull-back construction from fluid-matter 3-space has the same degrees of freedom as the number of independent components of the four-dimensional Lagrangian variations, supports the assumption. On the other hand, Andersson and Comer[1] argue that their variational model is compatible with non-vanishing particle creation rates. We support the result and prove that a proper variational formula considering the effect of the creation rate enables us to treat the non-vanishing creation rate without any additional requirement., Comment: 6 pages, No figure

We study the problem of heat conduction in general relativity by using Carter's variational formulation. We write the creation rates of the entropy and the particle as combinations of the vorticities of temperature and chemical potential. We pay attention to the fact that there are two additional degrees of freedom in choosing the relativistic analog of Cattaneo equation for the parts binormal to the caloric and the number flows. Including the contributions from the binormal parts, we find a $\textit{new}$ heat-flow equations and discover their dynamical role in thermodynamic systems. The benefit of introducing the binormal parts is that it allows room for a physical ansatz for describing the whole evolution of the thermodynamic system. Taking advantage of this platform, we propose a proper ansatz that deals with the binormal contributions starting from the physical properties of thermal equilibrium systems. We also consider the stability of a thermodynamic system in a flat background. We find that $\textit{new}$ "Klein" modes exist in addition to the known ones. We also find that the stability requirement is less stringent than those in the literature., Comment: 19 pages, 1 figure

Finding d-electron heavy fermion (HF) states has been an important topic as the diversity in d-electron materials can lead to many exotic Kondo effect-related phenomena or new states of matter such as correlation-driven topological Kondo insulator or cooperation between long-range magnetism and Kondo lattice behavior. Yet, obtaining direct spectroscopic evidence for a d-electron HF system has been elusive to date. Here, we report the observation of Kondo lattice behavior in an antiferromagnetic metal, FeTe, via angle-resolved photoemission spectroscopy (ARPES) and transport properties measurements. The Kondo lattice behavior is represented by the emergence of a sharp quasiparticle at low temperatures. The transport property measurements confirm the low-temperature Fermi liquid behavior and reveal successive coherent-incoherent crossover upon increasing temperature. We interpret the Kondo lattice behavior as a result of hybridization between localized Fe 3dxy and itinerant Te 5pz orbitals. Our interpretation is further evidenced by Fano-type tunneling spectra which accompany a hybridization gap. Our observations strongly suggest unusual cooperation between Kondo lattice behavior and long-range magnetic order., Comment: 17 pages, 6 figures

Understanding ultrafast magnetization dynamics on the microscopic level is of strong current interest due to the potential for applications in information storage. In recent years, the spin-lattice coupling has been recognized to be essential for ultrafast magnetization dynamics. Magnetoelectric multiferroics of type II possess intrinsic correlations among magnetic sublattices and electric polarization (P) through spin-lattice coupling, enabling fundamentally coupled dynamics between spins and lattice. Here we report on ultrafast magnetization dynamics in a room-temperature multiferroic hexaferrite possessing ferrimagnetic and antiferromagnetic sublattices, revealed by time-resolved resonant x-ray diffraction. A femtosecond above-bandgap excitation triggers a coherent magnon in which the two magnetic sublattices entangle and give rise to a transient modulation of P. A novel microscopic mechanism for triggering the coherent magnon in this ferrimagnetic insulator based on the spin-lattice coupling is proposed. Our finding opens up a novel but general pathway for ultrafast control of magnetism., Comment: 18 pages, 7 figures + 21 pages supplementary information

We study a rotating black hole with anisotropic matter and electromagnetic fields. We show that the electromagnetic field has the same form as the corresponding one in the Kerr-Newman geometry. The reason is thanks to the specific form of the metric function. We also mention energy extraction mechanisms from the black hole., Comment: 14 pages, Typo fixed. This article is prepared for the proceedings of the 17th Italian-Korean Symposium for Relativistic Astrophysics