Still searching Arxiv, refresh for possibly better results.

Results for "Zhifeng Zhang"

total 39034took 0.23s
Evaluation Mechanism of Collective Intelligence for Heterogeneous Agents GroupMar 01 2019May 28 2019Collective intelligence is manifested when multiple agents coherently work in observation, interaction, decision-making and action. In this paper, we define and quantify the intelligence level of heterogeneous agents group with the improved Anytime Universal ... More
Not Call Me Cellular Any More: The Emergence of Scaling Law, Fractal Patterns and Small-World in Wireless NetworksDec 02 2016Dec 15 2016In conventional cellular networks, for base stations (BSs) that are deployed far away from each other, it is general to assume them to be mutually independent. Nevertheless, after long-term evolution of cellular networks in various generations, this assumption ... More
Low-Rank Mechanism: Optimizing Batch Queries under Differential PrivacyAug 01 2012Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result, such that it is provably hard for the adversary to infer the presence or absence ... More
Topological Effect of Surface Plasmon Excitation in Gapped Isotropic Topological Insulator NanowiresSep 24 2013We present a theoretical investigation of the surface plasmon (SP) at the interface between topologically non-trivial cylindrical core and topological-trivial surrounding material, from the axion electrodynamics and modified constitutive relations. We ... More
Soil and soil CO2 magnify greenhouse effectJul 13 2019Soil has been recognized as an indirect driver of global warming by regulating atmospheric greenhouse gases. However, in view of the higher heat capacity and CO2 concentration in soil than those in atmosphere, the direct contributions of soil to greenhouse ... More
Supersymmetric microring laser arraysFeb 08 2019Feb 28 2019Coherent combination of emission power from an array of coupled semiconductor lasers operating on the same chip is of fundamental and technological importance. In general, the nonlinear competition among the array supermodes can entail incoherence and ... More
LibriTTS: A Corpus Derived from LibriSpeech for Text-to-SpeechApr 05 2019This paper introduces a new speech corpus called "LibriTTS" designed for text-to-speech use. It is derived from the original audio and text materials of the LibriSpeech corpus, which has been used for training and evaluating automatic speech recognition ... More
A Comparison of Constitutive Models of BloodAug 24 2018Mathematical models that accurately predict the mechanical behavior of blood can contribute to the development of biomedical devices and medications which are relevant in clinical applications. The models existing in the literature are complex enough ... More
Entry effects of droplet in a micro confinement: implications for deformation-based CTC microfiltrationJan 08 2016Deformation based circulating tumor cell (CTC) microchips are a representative diagnostic device for early cancer detection. This type of device usually involves a process of CTC trapping in a confined microgeometry. Further understanding of the CTC flow ... More
SADA: A General Framework to Support Robust Causation Discovery with Theoretical GuaranteeJul 05 2017Causation discovery without manipulation is considered a crucial problem to a variety of applications. The state-of-the-art solutions are applicable only when large numbers of samples are available or the problem domain is sufficiently small. Motivated ... More
Structural and Topological Nature of Plasticity in Sheared Granular MaterialsJun 25 2018Upon mechanical loading, granular materials yield and undergo plastic deformation. The nature of plastic deformation is essential for the development of the macroscopic constitutive models and the understanding of shear band formation. However, we still ... More
Hierarchical Generative Modeling for Controllable Speech SynthesisOct 16 2018Dec 27 2018This paper proposes a neural sequence-to-sequence text-to-speech (TTS) model which can control latent attributes in the generated speech that are rarely annotated in the training data, such as speaking style, accent, background noise, and recording conditions. ... More
Origin of Non-cubic Scaling Law in Disordered Granular PackingMay 19 2017Recent diffraction experiments on metallic glasses have unveiled an unexpected non-cubic scaling law between density and average interatomic distance, which lead to the speculations on the presence of fractal glass order. Using X-ray tomography we identify ... More
Gmail Smart Compose: Real-Time Assisted WritingMay 17 2019In this paper, we present Smart Compose, a novel system for generating interactive, real-time suggestions in Gmail that assists users in writing mails by reducing repetitive typing. In the design and deployment of such a large-scale and complicated system, ... More
Causal Discovery with Cascade Nonlinear Additive Noise ModelsMay 23 2019Jun 03 2019Identification of causal direction between a causal-effect pair from observed data has recently attracted much attention. Various methods based on functional causal models have been proposed to solve this problem, by assuming the causal process satisfies ... More
Joint Face Detection and Alignment using Multi-task Cascaded Convolutional NetworksApr 11 2016Face detection and alignment in unconstrained environment are challenging due to various poses, illuminations and occlusions. Recent studies show that deep learning approaches can achieve impressive performance on these two tasks. In this paper, we propose ... More
Causal Discovery with Cascade Nonlinear Additive Noise ModelsMay 23 2019Identification of causal direction between a causal-effect pair from observed data has recently attracted much attention. Various methods based on functional causal models have been proposed to solve this problem, by assuming the causal process satisfies ... More
Not Call Me Cellular Any More: The Emergence of Scaling Law, Fractal Patterns and Small-World in Wireless NetworksDec 02 2016In conventional cellular networks, for base stations (BSs) that are deployed far away from each other, it is general to assume them to be mutually independent. Nevertheless, after long-term evolution of cellular networks in various generations, the assumption ... More
Evaluation Mechanism of Collective Intelligence for Heterogeneous Agents GroupMar 01 2019Collective intelligence is manifested when multiple agents coherently work in observation, interaction, decision-making and action. In this paper, we define and quantify the intelligence level of heterogeneous agents group with the improved Anytime Universal ... More
The Collective Intelligence for Advancing 6G CommunicationsApr 26 2019May 05 2019The fifth-generation cellular networks (5G) has boosted the unprecedented convergence between the information world and physical world. On the other hand, empowered with the enormous amount of data and information, artificial intelligence (AI) has been ... More
Deep Learning-based Intelligent Dual Connectivity for Mobility Management in Dense NetworkMay 30 2018Ultra-dense network deployment has been proposed as a key technique for achieving capacity goals in the fifth-generation (5G) mobile communication system. However, the deployment of smaller cells inevitably leads to more frequent handovers, thus making ... More
GAN-based Deep Distributional Reinforcement Learning for Resource Management in Network SlicingMay 10 2019Jun 21 2019Network slicing is a key technology in 5G communications system, which aims to dynamically and efficiently allocate resources for diversified services with distinct requirements over a common underlying physical infrastructure. Therein, demand-aware resource ... More
Brain-Inspired Stigmergy LearningNov 20 2018Stigmergy has proved its great superiority in terms of distributed control, robustness and adaptability, thus being regarded as an ideal solution for large-scale swarm control problems. Based on new discoveries on astrocytes in regulating synaptic transmission ... More
The Collective Intelligence for Advancing 6G CommunicationsApr 26 2019May 12 2019The fifth-generation cellular networks (5G) has boosted the unprecedented convergence between the information world and physical world. On the other hand, empowered with the enormous amount of data and information, artificial intelligence (AI) has been ... More
AI-based Two-Stage Intrusion Detection for Software Defined IoT NetworksJun 07 2018Software Defined Internet of Things (SD-IoT) Networks profits from centralized management and interactive resource sharing which enhances the efficiency and scalability of IoT applications. But with the rapid growth in services and applications, it is ... More
On the Capacity of Fractal D2D Social Networks With Hierarchical CommunicationsAug 11 2017Aug 11 2018The maximum capacity of fractal D2D (device-to-device) social networks with both direct and hierarchical communications is studied in this paper. Specifically, the fractal networks are characterized by the direct social connection and the self-similarity. ... More
Convex Optimization for Linear Query Processing under Approximate Differential PrivacyFeb 13 2016May 16 2016Differential privacy enables organizations to collect accurate aggregates over sensitive data with strong, rigorous guarantees on individuals' privacy. Previous work has found that under differential privacy, computing multiple correlated aggregates as ... More
Deformability-based circulating tumor cell separation with conical-shaped microfilters: concept, optimization and design criteriaJan 08 2016The ability of detecting and separating CTCs can play a key role in early cancer detection and treatment. In recent years, there has been growing interest in using deformability-based CTC separation microfilters due to their simplicity and low cost. Most ... More
TACT: A Transfer Actor-Critic Learning Framework for Energy Saving in Cellular Radio Access NetworksNov 28 2012Apr 04 2014Recent works have validated the possibility of improving energy efficiency in radio access networks (RANs), achieved by dynamically turning on/off some base stations (BSs). In this paper, we extend the research over BS switching operations, which should ... More
On the Capacity of Fractal Wireless Networks With Direct Social InteractionsMay 27 2017The capacity of a fractal wireless network with direct social interactions is studied in this paper. Specifically, we mathematically formulate the self-similarity of a fractal wireless network by a power-law degree distribution $ P(k) $, and we capture ... More
Study on Base Station Topology in Cellular Networks: Take Advantage of Alpha Shapes, Betti Numbers, and Euler CharacteristicsAug 16 2018Faced with the ever-increasing trend of the cellular network scale, how to quantitatively evaluate the effectiveness of the large-scale deployment of base stations (BSs) has become a challenging topic. To this end, a deeper understanding of the cellular ... More
The Stochastic Geometry Analyses of Cellular Networks with α-Stable Self-SimilaritySep 18 2017Nov 19 2018To understand the spatial deployment of base stations (BSs) is the first step to analyze the performance of cellular networks and further design efficient networking protocols. Poisson point process (PPP), which has been widely adopted to characterize ... More
Fundamentals on Base Stations in Cellular Networks: From the Perspective of Algebraic TopologySep 18 2018In recent decades, the deployments of cellular networks have been going through an unprecedented expansion. In this regard, it is beneficial to acquire profound knowledge of cellular networks from the view of topology so that prominent network performances ... More
The Collective Intelligence for Advancing 6G CommunicationsApr 26 2019May 09 2019The fifth-generation cellular networks (5G) has boosted the unprecedented convergence between the information world and physical world. On the other hand, empowered with the enormous amount of data and information, artificial intelligence (AI) has been ... More
GAN-based Deep Distributional Reinforcement Learning for Resource Management in Network SlicingMay 10 2019Network slicing is a key technology in 5G communications system, which aims to dynamically and efficiently allocate resources for diversified services with distinct requirements over a common underlying physical infrastructure. Therein, demand-aware allocation ... More
Range Loss for Deep Face Recognition with Long-tailNov 28 2016Convolutional neural networks have achieved great improvement on face recognition in recent years because of its extraordinary ability in learning discriminative features of people with different identities. To train such a well-designed deep network, ... More
Deep Reinforcement Learning with Discrete Normalized Advantage Functions for Resource Management in Network SlicingJun 10 2019Network slicing promises to provision diversified services with distinct requirements in one infrastructure. Deep reinforcement learning (e.g., deep $\mathcal{Q}$-learning, DQL) is assumed to be an appropriate algorithm to solve the demand-aware inter-slice ... More
Multiplicity formula and stable trace formulaJul 30 2016Let $G$ be a connected reductive group over $\mathbb{Q}$. In this paper, we will stabilize the local trace formula, in particular, we construct the explicit form of the spectral side of stable local trace formula in the Archimedean case, when one component ... More
Multiplicity formula and stable trace formulaJul 30 2016Jun 05 2018Let $G$ be a connected reductive group over $\mathbb{Q}$. In this paper, we will stabilize the local trace formula, in particular, we construct the explicit form of the spectral side of stable local trace formula in the Archimedean case, when one component ... More
Trajectory-driven Influential Billboard PlacementFeb 06 2018In this paper we propose and study the problem of trajectory-driven influential billboard placement: given a set of billboards $U$ (each associated with a location and a cost), a database of trajectories $\mathcal{T}$ and a budget $L$, the goal is to ... More
Wave analysis in one dimensional structures with a wavelet finite element model and precise integration methodDec 05 2017Numerical simulation of ultrasonic wave propagation provides an efficient tool for crack identification in structures, while it requires a high resolution and expensive time calculation cost in both time integration and spatial discretization. Wavelet ... More
Trajectory-driven Influential Billboard PlacementFeb 06 2018Sep 15 2018In this paper we propose and study the problem of trajectory-driven influential billboard placement: given a set of billboards $U$ (each with a location and a cost), a database of trajectories $\mathcal{T}$ and a budget $L$, find a set of billboards within ... More
The Learning and Prediction of Application-level Traffic Data in Cellular NetworksJun 15 2016Traffic learning and prediction is at the heart of the evaluation of the performance of telecommunications networks and attracts a lot of attention in wired broadband networks. Now, benefiting from the big data in cellular networks, it becomes possible ... More
Combined effects of nonmetallic impurities and planned metallic dopants on grain boundary energy and strengthSep 06 2018Dec 12 2018Most research on nanocrystalline alloys has been focused on planned doping of metals with other metallic elements, but nonmetallic impurities are also prevalent in the real world. In this work, we report on the combined effects of metallic dopants and ... More
Characterizing Spatial Patterns of Base Stations in Cellular NetworksApr 04 2014Jun 19 2014The topology of base stations (BSs) in cellular networks, serving as a basis of networking performance analysis, is considered to be obviously distinctive with the traditional hexagonal grid or square lattice model, thus stimulating a fundamental rethinking. ... More
Two-tier Spatial Modeling of Base Stations in Cellular NetworksApr 04 2014Aug 01 2014Poisson Point Process (PPP) has been widely adopted as an efficient model for the spatial distribution of base stations (BSs) in cellular networks. However, real BSs deployment are rarely completely random, due to environmental impact on actual site planning. ... More
Traffic Prediction Based on Random Connectivity in Deep Learning with Long Short-Term MemoryNov 08 2017Apr 03 2018Traffic prediction plays an important role in evaluating the performance of telecommunication networks and attracts intense research interests. A significant number of algorithms and models have been put forward to analyse traffic data and make prediction. ... More
Effect of dynamical interactions on integrated properties of globular clustersJul 11 2014Nov 12 2014Globular Clusters (GCs) are generally treated as natural validators of simple stellar population (SSP) models. However, there are still some differences between real GCs and SSPs. In this work we use a direct $N$-body simulation code {\hf Nbody6} to study ... More
Efficient Decision-based Black-box Adversarial Attacks on Face RecognitionApr 09 2019Face recognition has obtained remarkable progress in recent years due to the great improvement of deep convolutional neural networks (CNNs). However, deep CNNs are vulnerable to adversarial examples, which can cause fateful consequences in real-world ... More
The Learning and Prediction of Application-level Traffic Data in Cellular NetworksJun 15 2016Mar 28 2017Traffic learning and prediction is at the heart of the evaluation of the performance of telecommunications networks and attracts a lot of attention in wired broadband networks. Now, benefiting from the big data in cellular networks, it becomes possible ... More
Convergence Analysis of the Dynamics of a Special Kind of Two-Layered Neural Networks with $\ell_1$ and $\ell_2$ RegularizationNov 19 2017In this paper, we made an extension to the convergence analysis of the dynamics of two-layered bias-free networks with one $ReLU$ output. We took into consideration two popular regularization terms: the $\ell_1$ and $\ell_2$ norm of the parameter vector ... More
State Transition Analysis of Time-Frequency Resource Conversion-based Call Admission Control for LTE-Type Cellular NetworkDec 02 2013To address network congestion stemmed from traffic generated by advanced user equipments, in [1] we propose a novel network resource allocation strategy, time-frequency resource conversion (TFRC), via exploiting user behavior, a specific kind of context ... More
Null Zig-Zag Wilson Loops in N=4 SYMMay 07 2009In planar ${\cal N}=4$ supersymmetric Yang-Mills theory we have studied supersymmetric Wilson loops composed of a large number of light-like segments, i.e., null zig-zags. These contours oscillate around smooth underlying spacelike paths. At one-loop ... More
LQCD at non zero isospin chemical potentialNov 02 2012Systems of non-zero isospin chemical potential are studied from a canonical approach by computing correlation functions with the quantum numbers of $N \pi^+$'s ($C_{N \pi}$). In order to reduce the number of contractions required in calculating $C_{N ... More
Occlusion Robust Face Recognition Based on Mask Learning with PairwiseDifferential Siamese NetworkAug 17 2019Deep Convolutional Neural Networks (CNNs) have been pushing the frontier of the face recognition research in the past years. However, existing general CNN face models generalize poorly to the scenario of occlusions on variable facial areas. Inspired by ... More
Multi-Tenant Cross-Slice Resource Orchestration: A Deep Reinforcement Learning ApproachJul 17 2018In a software-defined radio access network (RAN), a major challenge lies in how to support diverse services for mobile users (MUs) over a common physical network infrastructure. Network slicing is a promising solution to tailor the network to match such ... More
Phase Diagram and Magnetic Excitations of Anisotropic Spin-One MagnetsJan 10 2013We use a generalized spin wave approach and large scale quantum Monte Carlo (QMC) simulations to study the quantum phase diagram and quasiparticle excitations of the S=1 Heisenberg model with an easy-plane single-ion anisotropy in dimensions d=2 and 3. ... More
Multi-Tenant Cross-Slice Resource Orchestration: A Deep Reinforcement Learning ApproachJul 17 2018Jun 03 2019With the cellular networks becoming increasingly agile, a major challenge lies in how to support diverse services for mobile users (MUs) over a common physical network infrastructure. Network slicing is a promising solution to tailor the network to match ... More
Deep Learning with Long Short-Term Memory for Time Series PredictionOct 24 2018Time series prediction can be generalized as a process that extracts useful information from historical records and then determines future values. Learning long-range dependencies that are embedded in time series is often an obstacle for most algorithms, ... More
Optimizing Batch Linear Queries under Exact and Approximate Differential PrivacyFeb 26 2015Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result, such that it is provably hard for the adversary to infer the presence or absence ... More
Low Rank Mechanism for Optimizing Batch Queries under Differential PrivacyDec 11 2012Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result, such that it is provably hard for the adversary to infer the presence or absence ... More
Uncovering the influence of common nonmetal impurities on the stability and strength of a Σ5 (310) grain boundary in CuJun 18 2017Jan 30 2018Impurities are often driven to segregate to grain boundaries, which can significantly alter a material's thermal stability and mechanical behavior. To provide a comprehensive picture of this issue, the influence of a wide variety of common nonmetal impurities ... More
Orthogonal Deep Features Decomposition for Age-Invariant Face RecognitionOct 17 2018As facial appearance is subject to significant intra-class variations caused by the aging process over time, age-invariant face recognition (AIFR) remains a major challenge in face recognition community. To reduce the intra-class discrepancy caused by ... More
Large-scale Spatial Distribution Identification of Base Stations in Cellular NetworksNov 10 2014The performance of cellular system significantly depends on its network topology, where the spatial deployment of base stations (BSs) plays a key role in the downlink scenario. Moreover, cellular networks are undergoing a heterogeneous evolution, which ... More
Granular materials flow like complex fluidsNov 10 2017Granular materials such as sand, powders, foams etc. are ubiquitous in our daily life, as well as in industrial and geotechnical applications. Although these disordered systems form stable structures if unperturbed, in practice they do relax because of ... More
Translational and rotational dynamical heterogeneities in granular systemsMar 29 2018We use X-ray tomography to investigate the translational and rotational dynamical heterogeneities of a three dimensional hard ellipsoids granular packing driven by oscillatory shear. We find that particles which translate quickly form clusters with a ... More
Learning to Speak Fluently in a Foreign Language: Multilingual Speech Synthesis and Cross-Language Voice CloningJul 09 2019Jul 24 2019We present a multispeaker, multilingual text-to-speech (TTS) synthesis model based on Tacotron that is able to produce high quality speech in multiple languages. Moreover, the model is able to transfer voices across languages, e.g. synthesize fluent Spanish ... More
Learning to Speak Fluently in a Foreign Language: Multilingual Speech Synthesis and Cross-Language Voice CloningJul 09 2019We present a multispeaker, multilingual text-to-speech (TTS) synthesis model based on Tacotron that is able to produce high quality speech in multiple languages. Moreover, the model is able to transfer voices across languages, e.g. synthesize fluent Spanish ... More
Deep Reinforcement Learning for Resource Management in Network SlicingMay 17 2018Nov 21 2018Network slicing is born as an emerging business to operators, by allowing them to sell the customized slices to various tenants at different prices. In order to provide better-performing and cost-efficient services, network slicing involves challenging ... More
Experimental Realization of Multiple Topological Edge States in a One-Dimensional Photonic LatticeDec 13 2018Topological photonic systems offer light transport that is robust against defects and disorder, promising a new generation of chip-scale photonic devices and facilitating energy-efficient on-chip information routing and processing. However, present quasi ... More
A hybrid evolutionary algorithm with importance sampling for multi-dimensional optimizationAug 23 2013A hybrid evolutionary algorithm with importance sampling method is proposed for multi-dimensional optimization problems in this paper. In order to make use of the information provided in the search process, a set of visited solutions is selected to give ... More
Global existence and exponential decay of the solution for a viscoelastic wave equation with a delayNov 06 2012Apr 19 2013In this paper, we consider initial-boundary value problem of viscoelastic wave equation with a delay term in the interior feedback. Namely, we study the following equation $$u_{tt}(x,t)-\Delta u(x,t)+\int_0^t g(t-s)\Delta u(x,s)ds +\mu_1 u_t(x,t)+ \mu_2 ... More
Phonon Drag Effect in Nanocomposite FeSb2Oct 10 2012We study the temperature dependence of thermoelectric transport properties of four FeSb2 nanocomposite samples with different grain sizes. The comparison of the single crystals and nanocomposites of varying grain size indicates the presence of substantial ... More
Why Do Neural Response Generation Models Prefer Universal Replies?Aug 28 2018Recent advances in sequence-to-sequence learning reveal a purely data-driven approach to the response generation task. Despite its diverse applications, existing neural models are prone to producing short and generic replies, making it infeasible to tackle ... More
Three Dimensional Edwards-Anderson Spin Glass Model in an External FieldMar 18 2014We study the Edwards-Anderson model on a simple cubic lattice with a finite constant external field. We employ an indicator composed of a ratio of susceptibilities at finite wavenumbers, which was recently proposed to avoid the difficulties of a zero ... More
Optimizing thermoelectric performances of low-temperature SnSe compounds by electronic structure designApr 08 2015Recently SnSe compound was reported to have a peak thermoelectric figure-5 of-merit (ZT) of 2.62 at 923 K, but the ZT values at temperatures below 750 K are relatively low. In this work, the electronic structures of SnSe are calculated using the density ... More
Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence ModelsDec 05 2017Sequence-to-sequence models, such as attention-based models in automatic speech recognition (ASR), are typically trained to optimize the cross-entropy criterion which corresponds to improving the log-likelihood of the data. However, system performance ... More
Direct speech-to-speech translation with a sequence-to-sequence modelApr 12 2019Jun 25 2019We present an attention-based sequence-to-sequence neural network which can directly translate speech from one language into speech in another language, without relying on an intermediate text representation. The network is trained end-to-end, learning ... More
Supersymmetric microring laser arraysFeb 08 2019Coherent combination of emission power from an array of coupled semiconductor lasers operating on the same chip is of fundamental and technological importance. In general, the nonlinear competition among the array supermodes can entail incoherence and ... More
Metastable state involved resonant tunneling through single InAs/GaAs quantum dotAug 17 2007A scheme of resonant tunneling through the metastable state of semiconductor quantum dot is presented and implemented in the transport study of freestanding InAs quantum dots grown on GaAs(001) under illumination using conductive atomic force microscopy. ... More
A parallel-in-time multigrid solver with a new two-level convergence for two-dimensional unsteady fractional Laplacian problemsJun 17 2019The multigrid-reduction-in-time (MGRIT) technique has proven to be successful in achieving higher run-time speedup by exploiting parallelism in time. The goal of this article is to develop and analyze an MGRIT algorithm, using FCF-relaxation with time-dependent ... More
Improving the Performance of Online Neural Transducer ModelsDec 05 2017Having a sequence-to-sequence model which can operate in an online fashion is important for streaming applications such as Voice Search. Neural transducer is a streaming sequence-to-sequence model, but has shown a significant degradation in performance ... More
Existence of Dyons in Minimally Gauged Skyrme Model via Constrained MinimizationJul 01 2011We prove the existence of electrically and magnetically charged particlelike static solutions, known as dyons, in the minimally gauged Skyrme model developed by Brihaye, Hartmann, and Tchrakian. The solutions are spherically symmetric, depend on two continuous ... More
All-Silicon Topological Semimetals with Closed Nodal LineOct 28 2018Owing to the natural compatibility with current semiconductor industry, silicon allotropes with diverse structural and electronic properties provide promising platforms for the next-generation Si-based devices. After screening 230 all-silicon crystals ... More
Poisson Structures for Dispersionless Integrable Systems and Associated W-AlgebrasDec 04 1996Jan 13 1997In analogy to the KP theory, the second Poisson structure for the dispersionless KP hierarchy can be defined on the space of commutative pseudodifferential operators $L=p^n+\sum_{j=-\infty}^{n-1}u_j p^j$. The reduction of the Poisson structure to the ... More
Investigations of QCD at non-zero isospin densityNov 07 2011We investigate the QCD phase diagram as a function of isospin chemical potential at a fixed temperature by directly putting large numbers of \pi^+s into the system. Correlation functions of N \pi^+s systems involves N!N! contractions, and become extremely ... More
Generative Adversarial Networks with Inverse Transformation UnitSep 27 2017In this paper we introduce a new structure to Generative Adversarial Networks by adding an inverse transformation unit behind the generator. We present two theorems to claim the convergence of the model, and two conjectures to nonideal situations when ... More
Multi-Dialect Speech Recognition With A Single Sequence-To-Sequence ModelDec 05 2017Sequence-to-sequence models provide a simple and elegant solution for building speech recognition systems by folding separate components of a typical system, namely acoustic (AM), pronunciation (PM) and language (LM) models into a single neural network. ... More
Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech SynthesisJun 12 2018Jan 02 2019We describe a neural network-based system for text-to-speech (TTS) synthesis that is able to generate speech audio in the voice of many different speakers, including those unseen during training. Our system consists of three independently trained components: ... More
GradiVeQ: Vector Quantization for Bandwidth-Efficient Gradient Aggregation in Distributed CNN TrainingNov 08 2018Dec 31 2018Data parallelism can boost the training speed of convolutional neural networks (CNN), but could suffer from significant communication costs caused by gradient aggregation. To alleviate this problem, several scalar quantization techniques have been developed ... More
The spectral side of stable local trace formula for real groupsAug 30 2017Dec 04 2018Let $G$ be a connected quasi-split reductive group over $\mathbb{R}$, and more generally, a quasi-split $K$-group over $\mathbb{R}$. Arthur had obtained the formal formula for the spectral side of the stable local trace formula, by using formal substitute ... More
Comment on "Topological Nodal-Net Semimetal in a Graphene Network Structure"May 21 2019May 24 2019Recently, a distinct topological semimetal, nodal-net semimetal, has been identified by Wang et al. through ab initio calculations [Phys. Rev. Lett. 120, 026402 (2018)]. The authors claimed that a new body-centered tetragonal carbon allotrope with I4/mmm ... More
Boundary charges and integral identities for solitons in $(d+1)$-dimensional field theoriesOct 09 2017Nov 01 2017We establish a 3-parameter family of integral identities to be used on a class of theories possessing solitons with spherical symmetry in $d$ spatial dimensions. The construction provides five boundary charges that are related to certain integrals of ... More
Quarkonium at non-zero isospin densityNov 13 2012Apr 26 2013We calculate the energies of quarkonium bound states in the presence of a medium of nonzero isospin density using lattice QCD. The medium, created using a canonical (fixed isospin charge) approach, induces a reduction of the quarkonium energies. As the ... More
Self doping effect and successive magnetic transitions in superconducting Sr$_2$VFeAsO$_3$Jul 22 2010Sep 27 2010We have studied a quinary Fe-based superconductor Sr$_2$VFeAsO$_3$ by the measurements of x-ray diffraction, x-ray absorption, M\"{o}ssbauer spectrum, resistivity, magnetization and specific heat. This apparently undoped oxyarsenide is shown to be self ... More
Comment on "Topological Nodal-Net Semimetal in a Graphene Network Structure"May 21 2019Recently, a distinct topological semimetal, nodal-net semimetal, has been identified by Wang et al. through ab initio calculations [Phys. Rev. Lett. 120, 026402 (2018)]. The authors claimed that a new body-centered tetragonal carbon allotrope with I4/mmm ... More
Group Regularized Estimation under Structural HierarchyNov 17 2014Nov 08 2016Variable selection for models including interactions between explanatory variables often needs to obey certain hierarchical constraints. The weak or strong structural hierarchy requires that the existence of an interaction term implies at least one or ... More
The Best of Both Worlds: Combining Recent Advances in Neural Machine TranslationApr 26 2018Apr 27 2018The past year has witnessed rapid advances in sequence-to-sequence (seq2seq) modeling for Machine Translation (MT). The classic RNN-based approaches to MT were first out-performed by the convolutional seq2seq model, which was then out-performed by the ... More
Integer-Squared Laws for Global Vortices in the Born--Infeld Wave EquationsMay 07 2018Nov 11 2018A series of quantization identities are established for static vortex solutions governed by the Born--Infeld type actions. These identities are of a universal nature which are indifferent to the details of the models and provide refined descriptions of ... More
GPipe: Efficient Training of Giant Neural Networks using Pipeline ParallelismNov 16 2018Jul 25 2019Scaling up deep neural network capacity has been known as an effective approach to improving model quality for several different machine learning tasks. In many cases, increasing model capacity beyond the memory limit of a single accelerator has required ... More