arXiv subjects

Jing Zhou

Publications and source records attributed to Jing Zhou.

At least 145 records · Page 8Linked to original sources

arXiv2016

SYSTRAN's Pure Neural Machine Translation Systems

Since the first online demonstration of Neural Machine Translation (NMT) by LISA, NMT development has recently moved from laboratory to production systems as demonstrated by several entities announcing roll-out of NMT engines to replace their existing technologies. NMT systems have a large number of training configurations and the training process of such systems is usually very long, often a few weeks, so role of experimentation is critical and important to share. In this work, we present our approach to production-ready systems simultaneously with release of online demonstrators covering a large variety of languages (12 languages, for 32 language pairs). We explore different practical choices: an efficient and evolutive open-source framework; data preparation; network architecture; additional implemented features; tuning for production; etc. We discuss about evaluation methodology, present our first findings and we finally outline further work. Our ultimate goal is to share our expertise to build competitive production systems for "generic" translation. We aim at contributing to set up a collaborative framework to speed-up adoption of the technology, foster further research efforts and enable the delivery and adoption to/by industry of use-case specific engines integrated in real production workflows. Mastering of the technology would allow us to build translation engines suited for particular needs, outperforming current simplest/uniform systems.

cs.CL↗

arXiv2016

Beam energy dependence and updated test of the Trojan horse nucleus invariance via the d(d,p)t measurement at ultra-low energies

The $\mathrm{^2H}(d,p)\mathrm{^3H}$ bare nucleus astrophysical S(E) factor has been measured indirectly at energies from about 500 keV down to several keV by means of the Trojan-horse method applied to the quasi-free process $\mathrm{^2H({}^6Li},pt)\mathrm{^4He}$ induced at the lithium beam energy of 11 and 9.5 MeV, which makes the virtual binary process incident energy $\mathrm {E}_{dd}^{qf}$ go much closer to the zero-quasi-free-energy point than that in the previous similar experiment. The obtained results are compared with direct data as well as with previous indirect investigation of the same binary reactions. It shows that the precision of S(E) data in low energy range extracted via the same Trojan horse nucleus ($\mathrm{^6Li}=(d \oplus α)$ ) becomes better when the incident energy decreases from high value down to the zero-quasi-free-energy point. The very good agreement between data extracted from different Trojan horse nucleus ($\mathrm{^6Li}=(d \oplus α)$ vs. $\mathrm{^3He}=(d \oplus p)$) gives a strong updated test for the independence of the binary indirect cross section on the chosen Trojan horse nucleus at low energies.

nucl-ex↗

arXiv2016

Dark Matter Results from First 98.7-day Data of PandaX-II Experiment

We report the WIMP dark matter search results using the first physics-run data of the PandaX-II 500 kg liquid xenon dual-phase time-projection chamber, operating at the China JinPing Underground Laboratory. No dark matter candidate is identified above background. In combination with the data set during the commissioning run, with a total exposure of 3.3$\times10^4$ kg-day,the most stringent limit to the spin-independent interaction between the ordinary and WIMP dark matter is set for a range of dark matter mass between 3.5 and 1000 GeV/c$^2$. The best upper limit on the scattering cross section is found $2.5\times 10^{-46}$ cm$^2$ for the WIMP mass 40 GeV/c$^2$ at 90% confidence level.

hep-ex↗

arXiv2016

Dark Matter Search Results from the Commissioning Run of PandaX-II

We present the results of a search for WIMPs from the commissioning run of the PandaX-II experiment located at the China Jinping underground Laboratory. A WIMP search data set with an exposure of 306$\times$19.1 kg-day was taken, while its dominant $^{85}$Kr background was used as the electron recoil calibration. No WIMP candidates are identified, and a 90\% upper limit is set on the spin-independent elastic WIMP-nucleon cross section with a lowest excluded cross section of 2.97$\times$10$^{-45}$~cm$^2$ at a WIMP mass of 44.7~GeV/c$^2$.

hep-ex↗

arXiv2016

Recurrent Convolutional Neural Network Regression for Continuous Pain Intensity Estimation in Video

Automatic pain intensity estimation possesses a significant position in healthcare and medical field. Traditional static methods prefer to extract features from frames separately in a video, which would result in unstable changes and peaks among adjacent frames. To overcome this problem, we propose a real-time regression framework based on the recurrent convolutional neural network for automatic frame-level pain intensity estimation. Given vector sequences of AAM-warped facial images, we used a sliding-window strategy to obtain fixed-length input samples for the recurrent network. We then carefully design the architecture of the recurrent network to output continuous-valued pain intensity. The proposed end-to-end pain intensity regression framework can predict the pain intensity of each frame by considering a sufficiently large historical frames while limiting the scale of the parameters within the model. Our method achieves promising results regarding both accuracy and running speed on the published UNBC-McMaster Shoulder Pain Expression Archive Database.

cs.CV↗

arXiv2016

Exotic superfluid of trapped Fermi gases with spin-orbit coupling in dimensional crossover

We have studied the exotic superfluid phases of degenerated Fermi gases with spin-orbit coupling in a mixed dimensional system, where the motion of atoms are free in the x-direction and the tunneling between nearest tubes in the y-direction is permitted. Using the mean-field method, we obtain the phase diagrams of the system during the dimensional crossover between quasi one dimension to quasi two dimension. We find the existence of the topological state and Majorana edge mode in the weak tunneling case, and a rich phase diagram including two kinds of nodal superfluid phase and gapped superfluid phase in the opposite case. Our results show that topological pairing is favoured in quasi one dimension while nodal pairing state is favoured in quasi two dimension.

cond-mat.quant-gas↗

arXiv2015

Iso-oriented monolayer α-MoO3(010) films epitaxially grown on SrTiO3(001)

The ability to synthesis well-ordered two-dimensional materials under ultra-high vacuum and directly characterize them by other techniques in-situ can greatly advance our current understanding on their physical and chemical properties. In this paper, we demonstrate that iso-oriented α-MoO3 films with as low as single monolayer thickness can be reproducibly grown on SrTiO3(001) (STO) substrates by molecular beam epitaxy ( (010)MoO3 || (001)STO, [100]MoO3 || [100]STO or [010]STO) through a self-limiting process. While one in-plane lattice parameter of the MoO3 is very close to that of the SrTiO3 (aMoO3 = 3.96 Å, aSTO = 3.905 Å), the lattice mismatch along other direction is large (~5%, cMoO3 = 3.70 Å), which leads to relaxation as clearly observed from the splitting of streaks in reflection high-energy electron diffraction (RHEED) patterns. A narrow range in the growth temperature is found to be optimal for the growth of monolayer α-MoO3 films. Increasing deposition time will not lead to further increase in thickness, which is explained by a balance between deposition and thermal desorption due to the weak van der Waals force between α-MoO3 layers. Lowering growth temperature after the initial iso-oriented α-MoO3 monolayer leads to thicker α-MoO3(010) films with excellent crystallinity.

cond-mat.mtrl-sci↗

arXiv2015

Solutions for a nonlocal elliptic equation involving critical growth and Hardy potential

In this paper, by an approximating argument, we obtain infinitely many solutions for the following Hardy-Sobolev fractional equation with critical growth \begin{equation*}\label{0.1} \left\{% \begin{array}{ll} (-Δ)^{s} u-\ds\frac{μu}{|x|^{2s}}=|u|^{2^*_s-2}u+au, & \hbox{$\text{in}~ Ω$},\vspace{0.1cm} u=0,\,\, &\hbox{$\text{on}~\partial Ω$}, \\ \end{array}% \right. \end{equation*} provided $N>6s$, $μ\geq0$, $0< s<1$, $2^*_s=\frac{2N}{N-2s}$, $a>0$ is a constant and $Ω$ is an open bounded domain in $\R^N$ which contains the origin.

math.AP↗

arXiv2015

New measurement of the d(d,p)t reaction at astrophysical energies via the Trojan-horse method

The study of d(d,p)t reaction is very important for the nucleosynthesis in both standard Big Bang and stellar evolution, as well as for the future fusion reactors planning of energy production. The d(d,p)t bare nucleus astrophysical S(E) factor has been measured indirectly at energies from about 400 keV down to several keV by means of the Trojan horse method applied to the quasi-free process $\rm {}^2H({}^6Li,pt){}^4He$ induced at a lithium beam energy of 9.5 MeV, which is closer to the zero quasi-free energy point. An accurate analysis leads to the determination of the $\rm S_{bare}(0)=56.7 \pm 2.0 keV \cdot b$ and of the corresponding electron screening potential $\rm U_e = 13.2 \pm 4.3 eV$. In addition, this work gives an updated test for the Trojan horse nucleus invariance comparing with previous indirect investigations using $\rm {}^3He=(d+p)$ breakup.

nucl-ex↗

arXiv2014

Experimental spectra analysis in THM with the help of simulation based on Geant4 framework

The Coulomb barrier and electron screening cause difficulties in directly measuring nuclear reaction cross sections of charged particles in astrophysical energies. The Trojan-horse method has been introduced to solve the difficulties as a powerful indirect tool. In order to understand experimental spectra better, Geant4 is employed to simulate the method for the first time. Validity and reliability of the simulation are examined by comparing the experimental data with simulated results. The Geant4 simulation can give useful information to understand the experimental spectra better in data analysis and is beneficial to the design for future related experiments.

nucl-ex↗

arXiv2014

DAMIC at SNOLAB

We introduce the fully-depleted charge-coupled device (CCD) as a particle detector. We demonstrate its low energy threshold operation, capable of detecting ionizing energy depositions in a single pixel down to 50 eVee. We present results of energy calibrations from 0.3 keVee to 60 keVee, showing that the CCD is a fully active detector with uniform energy response throughout the silicon target, good resolution (Fano ~0.16), and remarkable linear response to electron energy depositions. We show the capability of the CCD to localize the depth of particle interactions within the silicon target. We discuss the mode of operation and unique imaging capabilities of the CCD, and how they may be exploited to characterize and suppress backgrounds. We present the first results from the deployment of 250 um thick CCDs in SNOLAB, a prototype for the upcoming DAMIC100. DAMIC100 will have a target mass of 0.1 kg and should be able to directly test the CDMS-Si signal within a year of operation.

physics.ins-det↗

arXiv2013

DAMIC: a novel dark matter experiment

DAMIC (Dark Matter in CCDs) is a novel dark matter experiment that has unique sensitivity to dark matter particles with masses below 10 GeV. Due to its low electronic readout noise (R.M.S. ~3 e-) this instrument is able to reach a detection threshold below 0.5 keV nuclear recoil energy, making the search for dark matter particles with low masses possible. We report on early results and experience gained from a detector that has been running at SNOLAB from Dec 2012. We also discuss the measured and expected backgrounds and present the plan for future detectors to be installed in 2014.

astro-ph.IM↗

arXiv2013

Bayesian factorizations of big sparse tensors

It has become routine to collect data that are structured as multiway arrays (tensors). There is an enormous literature on low rank and sparse matrix factorizations, but limited consideration of extensions to the tensor case in statistics. The most common low rank tensor factorization relies on parallel factor analysis (PARAFAC), which expresses a rank $k$ tensor as a sum of rank one tensors. When observations are only available for a tiny subset of the cells of a big tensor, the low rank assumption is not sufficient and PARAFAC has poor performance. We induce an additional layer of dimension reduction by allowing the effective rank to vary across dimensions of the table. For concreteness, we focus on a contingency table application. Taking a Bayesian approach, we place priors on terms in the factorization and develop an efficient Gibbs sampler for posterior computation. Theory is provided showing posterior concentration rates in high-dimensional settings, and the methods are shown to have excellent performance in simulations and several real data applications.

stat.ME↗

arXiv2013

Half-Metallic Silicene and Germanene Nanoribbons: towards High-Performance Spintronics Device

By using first-principles calculations, we predict that an in-plane homogenous electrical field can induce half-metallicity in hydrogen-terminated zigzag silicene and germanene nanoribbons (ZSiNRs and ZGeNRs). A dual-gated finite ZSiNR device reveals a nearly perfect spin-filter efficiency of up to 99% while a quadruple-gated finite ZSiNR device serves as an effective spin field effect transistor (FET) with an on/off current ratio of over 100 from ab initio quantum transport simulation. This discovery opens up novel prospect of silicene and germanene in spintronics.

cond-mat.mtrl-sci↗

arXiv2012

An Effective Method to Estimate Composition Amplitude of Spinodal Decomposition for Atom Probe Tomography Validated by Phase Field Simulations

Reasonable evaluation of composition amplitude in spinodal decomposition is a challenge to microanalysis of atom probe tomography, especially at early stages when phase separation is not prominent. This impedes quantitative analysis of spinodal structure in atom probe tomography as well as comparison with simulated results from phase field simulations. We hereby report an effective method to estimate the composition amplitude by constructing an amplitude density spectrum. This method can sensitively determine the composition amplitude at early stages. In particular, it substantially bridges experimental and simulation techniques comprising both discrete and continuum data in the study of spinodal decomposition. Moreover, it was found that the commonly adopted Langer-Bar-on-Miller method for atom probe analysis underestimates the composition amplitude of spinodal decomposition. Case studies have been performed on the Fe-Cr binary alloys.

cond-mat.mtrl-sci↗

arXiv2012

Solutions to the modified Korteweg-de Vries equation

This is a continuation of Ref.[1](arXiv:nlin.SI/0603008). In the present paper we review solutions to the modified Korteweg-de Vries equation in terms of Wronskians. The Wronskian entry vector needs to satisfy a matrix differential equation set which contains complex operation. This is different from the case of the Korteweg-de Vries equation. We introduce an auxiliary matrix to deal with the complex operation and then we are able to give complete solution expressions for the matrix differential equation set. The obtained solutions to the modified Korteweg-de Vries equation can simply be categorized by two types: solitons and breathers, together with their limit cases. Besides, we give rational solutions to the modified Korteweg-de Vries equation in Wromskian form. This is derived with the help of the Galilean transformed modified Korteweg-de Vries equation. Finally, typical dynamics of the obtained solutions is analyzed and illustrated. We list out the obtained solutions and their corresponding basic Wronskian vectors in the conclusion part.

nlin.SI↗

arXiv2011

Vortex structures of rotating spin-orbit coupled Bose-Einstein condensates

We consider the quasi-2D two-component Bose-Einstein condensates with Rashba spin-orbit (SO) coupling in a rotating trap. An external Zeeman term favoring spin polarization along the radial direction is also considered, which has the same form as the non-canonical part of the mechanical angular momentum. The rotating condensate exhibits rich structures as varying the strengths of trapping potential and interaction. With a strong trapping potential, the condensate exhibits a half-quantum vortex-lattice configuration. Such a configuration is driven to the normal one by introducing the external radial Zeeman field. In the case of a weak trap potential, the condensate exhibits a multi-domain pattern of plane-wave states under the external radial Zeeman field.

cond-mat.quant-gas↗

arXiv2011

Heating and Cooling in Adiabatic Mixing process

We study the effect of interaction on the temperature change in the process of adiabatic mixing two components of fermi gases by the real-space Bogoliubov-de Gennes (BdG) method. We find that in the process of adiabatic mixing, the competition of the adiabatic expansion and the attractive interaction make it possible to cool or heal the system depending on strength of interactions and the initial temperature. The change of temperature in a bulk system and a trapped system have been investigated respectively.

cond-mat.quant-gas↗