arXiv subjects

Liang Wu

Publications and source records attributed to Liang Wu.

At least 181 records · Page 10Linked to original sources

arXiv2019

Editing Text in the Wild

In this paper, we are interested in editing text in natural images, which aims to replace or modify a word in the source image with another one while maintaining its realistic look. This task is challenging, as the styles of both background and text need to be preserved so that the edited image is visually indistinguishable from the source image. Specifically, we propose an end-to-end trainable style retention network (SRNet) that consists of three modules: text conversion module, background inpainting module and fusion module. The text conversion module changes the text content of the source image into the target text while keeping the original text style. The background inpainting module erases the original text, and fills the text region with appropriate texture. The fusion module combines the information from the two former modules, and generates the edited text images. To our knowledge, this work is the first attempt to edit text in natural images at the word level. Both visual effects and quantitative results on synthetic and real-world dataset (ICDAR 2013) fully confirm the importance and necessity of modular decomposition. We also conduct extensive experiments to validate the usefulness of our method in various real-world applications such as text image synthesis, augmented reality (AR) translation, information hiding, etc.

cs.CV↗

arXiv2018

On the matter of topological insulators as magnetoelectrics

It has been proposed that topological insulators can be best characterized not as surface conductors, but as bulk magnetoelectrics that -- under the right conditions-- have a universal quantized magnetoelectric response coefficient $e^2/2h$. However, it is not clear to what extent these conditions are achievable in real materials that can have disorder, finite chemical potential, residual dissipation, and even inversion symmetry. This has led to some confusion and misconceptions. The primary goal of this work is to illustrate exactly under what real life scenarios and in what context topological insulators can be described as magnetoelectrics. We explore analogies of the 3D magnetoelectric response to electric polarization in 1D in detail, the formal vs. effective polarization and magnetoelectric susceptibility, the 1/2 quantized surface quantum Hall effect, the multivalued nature of the magnetoelectric susceptibility, the role of inversion symmetry, the effects of dissipation, and the necessity for finite frequency measurements. We present these issues from the perspective of experimentalists who have struggled to take the beautiful theoretical ideas and to try to measure their (sometimes subtle) physical consequences in messy real material systems.

cond-mat.mes-hall↗

arXiv2018

Field evolution of magnons in $α$-RuCl$_3$ by high-resolution polarized terahertz spectroscopy

The Kitaev quantum spin liquid (KSL) is a theoretically predicted state of matter whose fractionalized quasiparticles are distinct from bosonic magnons, the fundamental excitation in ordered magnets. The layered honeycomb antiferromagnet $α$-RuCl$_3$ is a KSL candidate material, as it can be driven to a magnetically disordered phase by application of an in-plane magnetic field, with $H_c \sim 7$ T. Here we report a detailed characterization of the magnetic excitation spectrum of this material by high-resolution time-domain terahertz (THz) spectroscopy. We observe two sharp magnon resonances whose frequencies and amplitudes exhibit a discontinuity as a function of applied magnetic field, as well as two broader peaks at higher energy. Below the Néel temperature, we find that linear spin wave theory can account for all of these essential features of the spectra when a $C_3$-breaking distortion of the honeycomb lattice and the presence of structural domains are taken into account.

cond-mat.str-el↗

arXiv2018

Multi-Level Network Embedding with Boosted Low-Rank Matrix Approximation

As opposed to manual feature engineering which is tedious and difficult to scale, network representation learning has attracted a surge of research interests as it automates the process of feature learning on graphs. The learned low-dimensional node vector representation is generalizable and eases the knowledge discovery process on graphs by enabling various off-the-shelf machine learning tools to be directly applied. Recent research has shown that the past decade of network embedding approaches either explicitly factorize a carefully designed matrix to obtain the low-dimensional node vector representation or are closely related to implicit matrix factorization, with the fundamental assumption that the factorized node connectivity matrix is low-rank. Nonetheless, the global low-rank assumption does not necessarily hold especially when the factorized matrix encodes complex node interactions, and the resultant single low-rank embedding matrix is insufficient to capture all the observed connectivity patterns. In this regard, we propose a novel multi-level network embedding framework BoostNE, which can learn multiple network embedding representations of different granularity from coarse to fine without imposing the prevalent global low-rank assumption. The proposed BoostNE method is also in line with the successful gradient boosting method in ensemble learning as multiple weak embeddings lead to a stronger and more effective one. We assess the effectiveness of the proposed BoostNE framework by comparing it with existing state-of-the-art network embedding methods on various datasets, and the experimental results corroborate the superiority of the proposed BoostNE network embedding framework.

cs.SI↗

arXiv2018

A Generalized Data Representation and Training-Performance Analysis for Deep Learning-Based Communications Systems

Deep learning (DL)-based autoencoder is a potential architecture to implement end-to-end communication systems. In this letter, we first give a brief introduction to the autoencoder-represented communication system. Then, we propose a novel generalized data representation (GDR) aiming to improve the data rate of DL-based communication systems. Finally, simulation results show that the proposed GDR scheme has lower training complexity, comparable block error rate performance and higher channel capacity than the conventional one-hot vector scheme. Furthermore, we investigate the effect of signal-to-noise ratio (SNR) in DL-based communication systems and prove that training at a high SNR could produce a good training performance for autoencoder.

cs.IT↗

arXiv2018

Large magneto-optical Kerr effect and imaging of magnetic octupole domains in an antiferromagnetic metal

When a polarized light beam is incident upon the surface of a magnetic material, the reflected light undergoes a polarization rotation. This magneto-optical Kerr effect (MOKE) has been intensively studied in a variety of ferro- and ferrimagnetic materials because it provides a powerful probe for electronic and magnetic properties as well as for various applications including magneto-optical recording. Recently, there has been a surge of interest in antiferromagnets (AFMs) as prospective spintronic materials for high-density and ultrafast memory devices, owing to their vanishingly small stray field and orders of magnitude faster spin dynamics compared to their ferromagnetic counterparts. In fact, the MOKE has proven useful for the study and application of the antiferromagnetic (AF) state. Although limited to insulators, certain types of AFMs are known to exhibit a large MOKE, as they are weak ferromagnets due to canting of the otherwise collinear spin structure. Here we report the first observation of a large MOKE signal in an AF metal at room temperature. In particular, we find that despite a vanishingly small magnetization of $M \sim$0.002 $μ_{\rm B}$/Mn, the non-collinear AF metal Mn$_3$Sn exhibits a large zero-field MOKE with a polar Kerr rotation angle of 20 milli-degrees, comparable to ferromagnetic metals. Our first-principles calculations have clarified that ferroic ordering of magnetic octupoles in the non-collinear Neel state may cause a large MOKE even in its fully compensated AF state without spin magnetization. This large MOKE further allows imaging of the magnetic octupole domains and their reversal induced by magnetic field. The observation of a large MOKE in an AF metal should open new avenues for the study of domain dynamics as well as spintronics using AFMs.

cond-mat.mtrl-sci↗

arXiv2018

Resonance-enhanced optical nonlinearity in the Weyl semimetal TaAs

While all media can exhibit first-order conductivity describing current linearly proportional to electric field, $E$, the second-order conductivity, $σ^{(2)}$ , relating current to $E^2$, is nonzero only when inversion symmetry is broken. Second order nonlinear optical responses are powerful tools in basic research, as probes of symmetry breaking, and in optical technology as the basis for generating currents from far-infrared to X-ray wavelengths. The recent surge of interest in Weyl semimetals with acentric crystal structures has led to the discovery of a host of $σ^{(2)}$ -related phenomena in this class of materials, such as polarization-selective conversion of light to dc current (photogalvanic effects) and the observation of giant second-harmonic generation (SHG) efficiency in TaAs at photon energy 1.5 eV. Here, we present measurements of the SHG spectrum of TaAs revealing that the response at 1.5 eV corresponds to the high-energy tail of a resonance at 0.7 eV, at which point the second harmonic conductivity is approximately 200 times larger than seen in the standard candle nonlinear crystal, GaAs. This remarkably large SHG response provokes the question of ultimate limits on $σ^{(2)}$ , which we address by a new theorem relating frequency-integrated nonlinear response functions to the third cumulant (or "skewness") of the polarization distribution function in the ground state. This theorem provides considerable insight into the factors that lead to the largest possible second-order nonlinear response, specifically showing that the spectral weight is unbounded and potentially divergent when the possibility of next-neighbor hopping is included.

cond-mat.mtrl-sci↗

arXiv2018

Nonvolatile Solid-State Charged-Polymer Gating of Topological Insulators into the Topological Insulating Regime

We demonstrate the ability to reduce the carrier concentration of thin films of the topological insulator (TI) Bi2Se3 by utilizing a novel approach, namely non-volatile electrostatic gating via corona charging of electret polymers. Sufficient electric field can be imparted to a polymer-TI bilayer to result in significant electron density depletion, even without the continuous connection of a gate electrode or the chemical modification of the TI. We show that the Fermi level of Bi2Se3 is shifted towards the Dirac point with this method. Using THz spectroscopy, we find that the surface chemical potential is lowered into the bulk band gap (~ 50 meV above the Dirac point and 170 meV below the conduction band minimum) and it is stabilized in the intrinsic regime while enhancing electron mobility. This represents the first use of a charged polymer gate for modulating TI charge density. The mobility of surface state electrons is enhanced to a value as high as ~1600 cm^2/Vs at 5K.

cond-mat.mes-hall↗

arXiv2018

A 3D Non-Stationary Wideband Geometry-Based Channel Model for MIMO Vehicle-to-Vehicle Communication System

In this paper, we present a three-dimensional (3D) non-wide-sense stationary (non-WSS) wideband geometry-based channel model for vehicle-to-vehicle (V2V) communication environments. We introduce a two-cylinder model to describe moving vehicles as well as multiple confocal semi-ellipsoid models to depict stationary roadside scenarios. The received signal is constructed as a sum of the line-of-sight (LoS), single-, and double-bounced rays with different energies. Accordingly, the proposed channel model is sufficient for depicting a wide variety of V2V environments, such as macro-, micro-, and picocells. The relative movement between the mobile transmitter (MT) and mobile receiver (MR) results in time-variant geometric statistics that make our channel model non-stationary. Using this channel model, the proposed channel statistics, i.e., the time-variant space correlation functions (CFs), frequency CFs, and corresponding Doppler power spectral density (PSD), were studied for different relative moving time instants. The numerical results demonstrate that the proposed 3D non-WSS wideband channel model is practical for characterizing real V2V channels.

eess.SP↗

arXiv2017

A compact broadband terahertz range quarter-wave plate

We detail the design and characterization of a terahertz range achromatic quarter-wave plate based on a stack of aligned variable thickness birefringent sapphire discs. The disc thicknesses and relative rotations of the discs are determined through a basin-hopping Monte Carlo thermal annealing routine. The basin-hopping scheme allows an improved refinement of the required thicknesses and rotations to give a predicted phase error from the ideal $π/2$ of only $0.5 \%$, which is a factor of approximately 6 better than previous efforts. Moreover, the large contrast between extraordinary and ordinary axes of sapphire allow us to greatly decrease the overall optical path length of our wave plate design by approximately a factor of 10 over similar designs based on quartz discs. However, this very same contrast requires very precise tolerances in the milled thicknesses of the discs and their assembly. We detail a method to compensate for differences in the thickness from their calculated ideal values. We have constructed one of our designs and found it similar in performance to other configurations, but with our much more compact geometry.

physics.ins-det↗

arXiv2017

Establishing and storing of deterministic quantum entanglement among three distant atomic ensembles

It is crucial for physical realization of quantum information networks to first establish entanglement among multiple space-separated quantum memories and then at a user-controlled moment to transfer the stored entanglement to quantum channels for distribution and conveyance of information. Here we present an experimental demonstration on generation, storage and transfer of deterministic quantum entanglement among three spatially separated atomic ensembles. The off-line prepared multipartite entanglement of optical modes is mapped into three distant atomic ensembles to establish entanglement of atomic spin waves via electromagnetically-induced-transparency light-matter interaction. Then the stored atomic entanglement is transferred into a tripartite quadrature entangled state of light, which is space-separated and can be dynamically allocated to three quantum channels for conveying quantum information. The existence of entanglement among released three optical modes verifies that the system has capacity of preserving multipartite entanglement. The presented protocol can be directly extended to larger quantum networks with more nodes.

quant-ph↗

arXiv2017

Experimental generation of tripartite polarization entangled states of bright optical beams

The multipartite polarization entangled states of bright optical beams directly associating with the spin states of atomic ensembles are one of the essential resources in the future quantum information networks, which can be conveniently utilized to transfer and convert quantum states across a network composed of many atomic nodes. In this letter, we present the experimental demonstration of tripartite polarization entanglement described by Stokes operators of optical field. The tripartite entangled states of light at the frequency resonant with D1 line of Rubidium atoms are transformed into the continuous variable polarization entanglement among three bright optical beams via an optical beam splitter network. The obtained entanglement is confirmed by the extended criterion for polarization entanglement of multipartite quantized optical modes.

quant-ph↗

arXiv2017

Antiferromagnetic resonance and terahertz continuum in $α-$RuCl$_3$

We report measurements of optical absorption in the zig-zag antiferromagnet $α$-RuCl$_3$ as a function of temperature, $T$, magnetic field, $B$, and photon energy, $\hbarω$ in the range $\sim$ 0.3 to 8.3 meV, using time-domain terahertz spectroscopy. Polarized measurements show that 3-fold rotational symmetry is broken in the honeycomb plane from 2 K to 300 K. We find a sharp absorption peak at 2.56 meV upon cooling below the Néel temperature of 7 K at $B=0$ that we identify as magnetic-dipole excitation of a zero-wavevector magnon, or antiferromagnetic resonance (AFMR). With application of $B$, the AFMR broadens and shifts to lower frequency as long-range magnetic order is lost in a manner consistent with transitioning to a spin-disordered phase. From direct, internally calibrated measurement of the AFMR spectral weight, we place an upper bound on the contribution to the $dc$ susceptibility from a magnetic excitation continuum.

cond-mat.str-el↗

arXiv2017

Imaging anomalous nematic order and strain in optimally doped BaFe$_2$(As,P)$_2$

We present the strain and temperature dependence of an anomalous nematic phase in optimally doped BaFe$_2$(As,P)$_2$. Polarized ultrafast optical measurements reveal broken 4-fold rotational symmetry in a temperature range above $T_c$ in which bulk probes do not detect a phase transition. Using ultrafast microscopy, we find that the magnitude and sign of this nematicity vary on a ${50{-}100}~μ$m length scale, and the temperature at which it onsets ranges from 40 K near a domain boundary to 60 K deep within a domain. Scanning Laue microdiffraction maps of local strain at room temperature indicate that the nematic order appears most strongly in regions of weak, isotropic strain. These results indicate that nematic order arises in a genuine phase transition rather than by enhancement of local anisotropy by a strong nematic susceptibility. We interpret our results in the context of a proposed surface nematic phase.

cond-mat.supr-con↗

arXiv2017

Giant anisotropic nonlinear optical response in transition metal monopnictide Weyl semimetals

Although Weyl fermions have proven elusive in high-energy physics, their existence as emergent quasiparticles has been predicted in certain crystalline solids in which either inversion or time-reversal symmetry is broken\cite{WanPRB2011,BurkovPRL2011, WengPRX2015,HuangNatComm2015}. Recently they have been observed in transition metal monopnictides (TMMPs) such as TaAs, a class of noncentrosymmetric materials that heretofore received only limited attention \cite{XuScience2015, LvPRX2015, YangNatPhys2015}. The question that arises now is whether these materials will exhibit novel, enhanced, or technologically applicable electronic properties. The TMMPs are polar metals, a rare subset of inversion-breaking crystals that would allow spontaneous polarization, were it not screened by conduction electrons \cite{anderson1965symmetry,shi2013ferroelectric,kim2016polar}. Despite the absence of spontaneous polarization, polar metals can exhibit other signatures of inversion-symmetry breaking, most notably second-order nonlinear optical polarizability, $χ^{(2)}$, leading to phenomena such as optical rectification and second-harmonic generation (SHG). Here we report measurements of SHG that reveal a giant, anisotropic $χ^{(2)}$ in the TMMPs TaAs, TaP, and NbAs. With the fundamental and second harmonic fields oriented parallel to the polar axis, the value of $χ^{(2)}$ is larger by almost one order of magnitude than its value in the archetypal electro-optic materials GaAs \cite{bergfeld2003second} and ZnTe \cite{wagner1998dispersion}, and in fact larger than reported in any crystal to date.

cond-mat.mtrl-sci↗

arXiv2017

A Scalable Algorithm for Structure Identification of Complex Gene Regulatory Network from Temporal Expression Data

Motivation: Gene regulatory interactions are of fundamental importance to various biological functions and processes. However, only a few previous computational studies have claimed success in revealing genome-wide regulatory landscapes from temporal gene expression data, especially for complex eukaryotes like human. Moreover, recent work suggests that these methods still suffer from the curse of dimensionality if network size increases to 100 or higher. Result: We present a novel scalable algorithm for identifying genome-wide regulatory network structures. The highlight of our method is that its superior performance does not degenerate even for a network size on the order of $10^4$, and is thus readily applicable to large-scale complex networks. Such a breakthrough is achieved by considering both prior biological knowledge and multiple topological properties (i.e., sparsity and hub gene structure) of complex networks in the regularized formulation. We also illustrate the application of our algorithm in practice using the time-course expression data from an influenza infection study in respiratory epithelial cells. Availability and Implementation: The algorithm described in this article is implemented in MATLAB$^\circledR$. The source code is freely available from https://github.com/Hongyu-Miao/DMI.git. Contact: jliu@cs.rochester.edu; hongyu.miao@uth.tmc.edu Supplementary information: Supplementary data are available online.

q-bio.MN↗

arXiv2016

Quantized Faraday and Kerr rotation and axion electrodynamics of a 3D topological insulator

Topological insulators have been proposed to be best characterized as bulk magnetoelectric materials that show response functions quantized in terms of fundamental physical constants. Here we lower the chemical potential of three-dimensional (3D) Bi$_2$Se$_3$ films to $\sim$ 30 meV above the Dirac point, and probe their low-energy electrodynamic response in the presence of magnetic fields with high-precision time-domain terahertz polarimetry. For fields higher than 5 T, we observed quantized Faraday and Kerr rotations, whereas the DC transport is still semi-classical. A non-trivial Berry phase offset to these values gives evidence for axion electrodynamics and the topological magnetoelectric effect. The time structure used in these measurements allows a direct measure of the fine structure constant based on a topological invariant of a solid-state system.

cond-mat.mes-hall↗

arXiv2016

Disorder-driven topological phase transition in Bi2Se3 films

Topological insulators (TI) are a phase of matter that host unusual metallic states on their surfaces. Unlike the states that exist on the surface of conventional materials, these so-called topological surfaces states (TSS) are protected against disorder-related localization effects by time reversal symmetry through strong spin-orbit coupling. By combining transport measurements, angle-resolved photo-emission spectroscopy and scanning tunneling microscopy, we show that there exists a critical level of disorder beyond which the TI Bi2Se3 loses its ability to protect the metallic TSS and transitions to a fully insulating state. The absence of the metallic surface channels dictates that there is a change in topological character, implying that disorder can lead to a topological phase transition even without breaking the time reversal symmetry. This observation challenges the conventional notion of topologically-protected surface states, and will provoke new studies as to the fundamental nature of topological phase of matter in the presence of disorder.

cond-mat.mtrl-sci↗