arXiv subjects

Bin Xu

Publications and source records attributed to Bin Xu.

At least 163 records · Page 9Linked to original sources

arXiv2022

Solutions of the ${\rm SU}(n+1)$ Toda system from meromorphic functions

We consider the ${\rm SU}(n+1)$ Toda system on a simply connected domain $Ω$ in ${\Bbb C}$, the $n=1$ case of which coincides with the Liouville equation $Δu+8e^u=0$. A classical result by Liouville says that a solution of this equation on $Ω$ can be represented by some non-degenerate meromorphic function on $Ω$. We construct a family of solutions parameterized by ${\rm PSL}(n+1,\,{\Bbb C})/{\rm PSU}(n+1)$ for the ${\rm SU}(n+1)$ Toda system from such a meromorphic function on $Ω$, which generalizes the result of Liouville. As an application, we find a new class of solvable ${\rm SU}(n+1)$ Toda systems with singular sources via cone spherical metrics on compact Riemann surfaces.

math-ph↗

arXiv2022

A new look at the temperature-dependent properties of the antiferroelectric model PbZrO3: an effective Hamiltonian study

A novel atomistic effective Hamiltonian scheme, incorporating an original and simple bilinear energetic coupling, is developed and used to investigate the temperature dependent physical properties of the prototype antiferroelectric PbZrO3 (PZO) system. This scheme reproduces very well the known experimental hallmarks of the complex Pbam orthorhombic phase at low temperatures and the cubic paraelectric state of Pm 3m symmetry at high temperatures. Unexpectedly, it further predicts a novel intermediate state also of Pbam symmetry, but in which anti-phase oxygen octahedral tiltings have vanished with respect to the Pbam ground state. Interestingly, such new state exhibits a large dielectric response and thermal expansion that remarkably agree with previous experimental observations and the x-ray experiments we performed. We also conducted direct first-principles calculations at 0K which further support such low energy phase. Within this fresh framework, a re-examination of the properties of PZO is thus called for.

cond-mat.mtrl-sci↗

arXiv2022

Spin-current driven Dzyaloshinskii-Moriya interaction in the multiferroic BiFeO3 from first-principles

The electrical control of magnons opens up new ways to transport and process information for logic devices. In magnetoelectrical multiferroics, the Dzyaloshinskii-Moriya (DM) interaction directly allow for such a control and, hence, is of major importance. We determine the origin and the strength of the (converse) spin current DM interaction in the R3c bulk phase of the multiferroic BiFeO3 based on density functional theory. Our data supports only the existence of one DM interaction contribution originating from the spin current model. By exploring then magnon dispersion in the full Brillouin Zone, we show that the exchange is isotropic, but the DM interaction and anisotropy prefer any propagation and any magnetization direction within the full (111) plane. Our work emphasizes the significance of the asymmetric potential induced by the spin current over the structural asymmetry induced by the anionic octahedron in multiferroics such as BiFeO3.

cond-mat.mtrl-sci↗

arXiv2022

EDUKG: a Heterogeneous Sustainable K-12 Educational Knowledge Graph

Web and artificial intelligence technologies, especially semantic web and knowledge graph (KG), have recently raised significant attention in educational scenarios. Nevertheless, subject-specific KGs for K-12 education still lack sufficiency and sustainability from knowledge and data perspectives. To tackle these issues, we propose EDUKG, a heterogeneous sustainable K-12 Educational Knowledge Graph. We first design an interdisciplinary and fine-grained ontology for uniformly modeling knowledge and resource in K-12 education, where we define 635 classes, 445 object properties, and 1314 datatype properties in total. Guided by this ontology, we propose a flexible methodology for interactively extracting factual knowledge from textbooks. Furthermore, we establish a general mechanism based on our proposed generalized entity linking system for EDUKG's sustainable maintenance, which can dynamically index numerous heterogeneous resources and data with knowledge topics in EDUKG. We further evaluate EDUKG to illustrate its sufficiency, richness, and variability. We publish EDUKG with more than 252 million entities and 3.86 billion triplets. Our code and data repository is now available at https://github.com/THU-KEG/EDUKG.

cs.CL↗

arXiv2022

Origin of negative electrocaloric effect in Pnma-type antiferroelectric perovskites

Anomalous electrocaloric effect (ECE) with decreasing temperature upon application of an electric field is known to occur in antiferroelectrics (AFEs), and previous understanding refers to the field-induced canting of electric dipoles if there is no phase transitions. Here, we use a first-principle-based method to study the ECE in Nd-substituted BiFeO3 (BNFO) perovskite solid solutions, which has the Pnma-type AFE ground state. We demonstrate another scenario to achieve and explain anomalous ECE, emphasizing that explicit consideration of octahedral tiltings is indispensable for a correct understanding. This mechanism may be general for AFEs for which the antipolar mode is not the primary order parameter. We also find that the negative ECE can reach a large magnitude in BNFO.

cond-mat.mtrl-sci↗

arXiv2022

ConstGCN: Constrained Transmission-based Graph Convolutional Networks for Document-level Relation Extraction

Document-level relation extraction with graph neural networks faces a fundamental graph construction gap between training and inference - the golden graph structure only available during training, which causes that most methods adopt heuristic or syntactic rules to construct a prior graph as a pseudo proxy. In this paper, we propose $\textbf{ConstGCN}$, a novel graph convolutional network which performs knowledge-based information propagation between entities along with all specific relation spaces without any prior graph construction. Specifically, it updates the entity representation by aggregating information from all other entities along with each relation space, thus modeling the relation-aware spatial information. To control the information flow passing through the indeterminate relation spaces, we propose to constrain the propagation using transmitting scores learned from the Noise Contrastive Estimation between fact triples. Experimental results show that our method outperforms the previous state-of-the-art (SOTA) approaches on the DocRE dataset.

cs.CL↗

arXiv2022

Unusual slow energy relaxation induced by mobile discrete breathers in one-dimensional lattices with next-nearest-neighbor coupling

We study the energy relaxation process in one-dimensional (1D) lattices with next-nearest-neighbor (NNN) couplings. This relaxation is produced by adding damping (absorbing conditions) to the boundary (free-end) of the lattice. Compared to the 1D lattices with on-site potentials, the properties of discrete breathers (DBs) that are spatially localized intrinsic modes are quite unusual with the NNN couplings included, i.e., these DBs are mobile, and thus they can interact with both the phonons and the boundaries of the lattice. For the interparticle interactions of harmonic and Fermi-Pasta-Ulam-Tsingou-$β$ (FPUT-$β$) types, we find two crossovers of relaxation in general, i.e., a first crossover from the stretched-exponential to the regular exponential relaxation occurring in a short timescale, and a further crossover from the exponential to the power-law relaxation taking place in a long timescale. The first and second relaxations are universal, but the final power-law relaxation is strongly influenced by the properties of DBs, e.g. the scattering processes of DBs with phonons and boundaries in the FPUT-$β$ type systems make the power-law decay relatively faster than that in the counterparts of the harmonic type systems under the same coupling. Our results present new information and insights for understanding the slow energy relaxation in cooling the lattices.

nlin.CD↗

arXiv2022

Phase-transition-induced giant Thomson effect for thermoelectric cooling

The Seebeck and Peltier effects have been widely studied and used in various thermoelectric technologies, including thermal energy harvesting and solid-state heat pumps. However, basic and applied studies on the Thomson effect, another fundamental thermoelectric effect in conductors, are limited despite the fact that the Thomson effect allows electronic cooling through the application of a temperature gradient bias rather than the construction of junction structures. In this article, we report the observation of a giant Thomson effect that appears owing to magnetic phase transitions. The Thomson coefficient of FeRh-based alloys reaches large values approaching $-$1,000 $μ$VK$^{-1}$ around room temperature because of the steep temperature dependence of the Seebeck coefficient associated with the antiferromagnetic-ferromagnetic phase transition. The Thomson coefficient is several orders of magnitude larger than the Seebeck coefficient of the alloys. Using the active thermography technique, we demonstrate that the Thomson cooling can be much larger than Joule heating in the same material even in a nearly steady state. The operation temperature of the giant Thomson effect in the FeRh-based alloys can be tuned over a wide range by applying an external magnetic field or by slightly changing the composition. Our findings provide a new direction in the materials science of thermoelectrics and pave the way for thermal management applications using the Thomson effect.

cond-mat.mtrl-sci↗

arXiv2022

A Critical Analysis of Image-based Camera Pose Estimation Techniques

Camera, and associated with its objects within the field of view, localization could benefit many computer vision fields, such as autonomous driving, robot navigation, and augmented reality (AR). In this survey, we first introduce specific application areas and the evaluation metrics for camera localization pose according to different sub-tasks (learning-based 2D-2D task, feature-based 2D-3D task, and 3D-3D task). Then, we review common methods for structure-based camera pose estimation approaches, absolute pose regression and relative pose regression approaches by critically modelling the methods to inspire further improvements in their algorithms such as loss functions, neural network structures. Furthermore, we summarise what are the popular datasets used for camera localization and compare the quantitative and qualitative results of these methods with detailed performance metrics. Finally, we discuss future research possibilities and applications.

cs.CV↗

arXiv2021

OpenQA: Hybrid QA System Relying on Structured Knowledge Base as well as Non-structured Data

Search engines based on keyword retrieval can no longer adapt to the way of information acquisition in the era of intelligent Internet of Things due to the return of keyword related Internet pages. How to quickly, accurately and effectively obtain the information needed by users from massive Internet data has become one of the key issues urgently needed to be solved. We propose an intelligent question-answering system based on structured KB and unstructured data, called OpenQA, in which users can give query questions and the model can quickly give accurate answers back to users. We integrate KBQA structured question answering based on semantic parsing and deep representation learning, and two-stage unstructured question answering based on retrieval and neural machine reading comprehension into OpenQA, and return the final answer with the highest probability through the Transformer answer selection module in OpenQA. We carry out preliminary experiments on our constructed dataset, and the experimental results prove the effectiveness of the proposed intelligent question answering system. At the same time, the core technology of each module of OpenQA platform is still in the forefront of academic hot spots, and the theoretical essence and enrichment of OpenQA will be further explored based on these academic hot spots.

cs.CL↗

arXiv2021

3+1 Dimension Schwinger Pair Production with Quantum Computers

Real-time quantum simulation of quantum field theory in (3+1)D requires large quantum computing resources. With a few-qubit quantum computer, we develop a novel algorithm and experimentally study the Schwinger effect, the electron-positron pair production in a strong electric field, in (3+1)D. The resource reduction is achieved by treating the electric field as a background field, working in Fourier space transverse to the electric field direction, and considering parity symmetry, such that we successfully map the three spatial dimension problems into one spatial dimension problems. We observe that the rate of pair production of electrons and positrons is consistent with the theoretical predication of the Schwinger effect. Our work paves the way towards exploring quantum simulation of quantum field theory beyond one spatial dimension.

quant-ph↗

arXiv2021

Structural Origin of Boson Peak in Glasses

Boson peak, the excess low energy excitations in the terahertz regime, is one of the most unique features of disordered systems and has been linked to many anomalous properties of glass materials. The nature and structural origin of the boson peak remain elusive and have been debated for more than a half century mainly due to the lack of real-time and real-space experimental insights of the dynamic phenomenon. In this work we employed femtosecond MeV ultrafast electron diffraction to characterize the atomic dynamics of metallic glasses in real time. The experiment reveals collective atomic oscillations, presented in elastic electron scattering and atomic pair distribution functions, within the boson peak frequency range of 1.0-1.8 THz in both reciprocal and real space. It was found that the oscillation frequency has reciprocal dependence on interatomic pair distances and the corresponding wave velocity experimentally affirms the transverse acoustic wave nature of the boson peak. The observed strong correlation between THz acoustic vibrations and coherent electron scattering provides compelling evidence that the boson peak originates from the collective transverse vibrational modes of structurally ordered atoms in the disordered system.

cond-mat.dis-nn↗

arXiv2021

Design of an Adaptive Surface for Space-Reconfigurable Reflectors using Auxetic Lattice Skin

The effect of Poisson's ratio to the reflector reshaping is investigated through mechanical study of reconfigurable reflectors in this paper. The value of Poisson's ratio corresponding to the minimum deforming stress is given and an auxetic lattice is proposed for the reflector surface. The parameters of the auxetic lattice are investigated for vary Poisson's ratio. A case of reconfigurable reflector is studied, the curvature change and strain are calculated by surface geometry analysis, and the negative Poisson's ratio is established for vary thickness. According to RMS calculation by the FEM structure analysis, the thickness can finally be established.

cond-mat.soft↗

arXiv2021

A flexible film thermocouple temperature sensor

This article introduces a thin-film thermocouple temperature sensor with symmetrical electrode structure. It uses PI film as a flexible substrate. Cu film and CuNi film made by MEMS manufacturing process as positive and negative electrodes. The device itself has the advantages of miniature, bendable and fast response speed. To reduce the film resistance value. Conducting metal film thickness and sputtering substrate temperature optimization experiments. The critical dimensions of Cu/CuNi film are 650nm and 400nm. The best sputtering substrate temperature for Cu/CuNi films is 100 degrees Celsius and 150 degrees Celsius. Testing the adhesion of thin film thermocouples using the peel-off method. The test result is 9.4N. Finally, the film thermocouple temperature sensor is subjected to a temperature static calibration experiment. The result shows that the actual potential difference error is within 1 degrees Celsius. It belongs to the second class standard in the formulation of thermocouple standards in my country. Through curve fitting, the corresponding relationship between temperature and potential difference is more accurate.

cond-mat.mtrl-sci↗

arXiv2021

Geometry reconfiguration of a reflector with an auxetic material surface

Mechanically reconfigurable reflectors can modify their geometries to realize large angle beam scanning. This paper studies a method for reconfiguring a reflector geometry from a standard parabolic shape to offset parabolic shapes with varies offset angles. To fulfil the geometry control, the coordinate change of the nodes on the reflector surface are studied with the coordinate transformation method. To accommodate the large deformation of the reflector surface, a negative Poisson's ratio (NPR) material is applied to the reconfigurable reflector.

cond-mat.mtrl-sci↗

arXiv2021

A capacitive flexible tactile sensor

In this paper, a capacitive flexible tactile sensor was designed to measure the pressure of objects based on MEMS technology. This sensor is a structure of a 4x4 array, with metal Ag as the capacitive electrode, which forms the tactile sensing unit of the sensor. The structure of capacitive flexible tactile sensor was designed and an experimental platform was established to test the performance. The tests show that when the thickness of the intermediate layer is 2 mm and the density is medium, the sensor's sensitivity is the best while the time of both the response and the rebound is fast.

cond-mat.mtrl-sci↗

arXiv2021

Improving Low-resource Reading Comprehension via Cross-lingual Transposition Rethinking

Extractive Reading Comprehension (ERC) has made tremendous advances enabled by the availability of large-scale high-quality ERC training data. Despite of such rapid progress and widespread application, the datasets in languages other than high-resource languages such as English remain scarce. To address this issue, we propose a Cross-Lingual Transposition ReThinking (XLTT) model by modelling existing high-quality extractive reading comprehension datasets in a multilingual environment. To be specific, we present multilingual adaptive attention (MAA) to combine intra-attention and inter-attention to learn more general generalizable semantic and lexical knowledge from each pair of language families. Furthermore, to make full use of existing datasets, we adopt a new training framework to train our model by calculating task-level similarities between each existing dataset and target dataset. The experimental results show that our XLTT model surpasses six baselines on two multilingual ERC benchmarks, especially more effective for low-resource languages with 3.9 and 4.1 average improvement in F1 and EM, respectively.

cs.CL↗

arXiv2021

DiaKG: an Annotated Diabetes Dataset for Medical Knowledge Graph Construction

Knowledge Graph has been proven effective in modeling structured information and conceptual knowledge, especially in the medical domain. However, the lack of high-quality annotated corpora remains a crucial problem for advancing the research and applications on this task. In order to accelerate the research for domain-specific knowledge graphs in the medical domain, we introduce DiaKG, a high-quality Chinese dataset for Diabetes knowledge graph, which contains 22,050 entities and 6,890 relations in total. We implement recent typical methods for Named Entity Recognition and Relation Extraction as a benchmark to evaluate the proposed dataset thoroughly. Empirical results show that the DiaKG is challenging for most existing methods and further analysis is conducted to discuss future research direction for improvements. We hope the release of this dataset can assist the construction of diabetes knowledge graphs and facilitate AI-based applications.

cs.CL↗