arXiv subjects

Sheng Zhang

Publications and source records attributed to Sheng Zhang.

At least 253 records · Page 14Linked to original sources

arXiv2018

Cross-lingual Semantic Parsing

We introduce the task of cross-lingual semantic parsing: mapping content provided in a source language into a meaning representation based on a target language. We present: (1) a meaning representation designed to allow systems to target varying levels of structural complexity (shallow to deep analysis), (2) an evaluation metric to measure the similarity between system output and reference meaning representations, (3) an end-to-end model with a novel copy mechanism that supports intrasentential coreference, and (4) an evaluation dataset where experiments show our model outperforms strong baselines by at least 1.18 F1 score.

cs.CL↗

arXiv2018

FPGA Implementations of 3D-SIMD Processor Architecture for Deep Neural Networks Using Relative Indexed Compressed Sparse Filter Encoding Format and Stacked Filters Stationary Flow

It is a challenging task to deploy computationally and memory intensive State-of-the-art deep neural networks (DNNs) on embedded systems with limited hardware resources and power budgets. Recently developed techniques like Deep Compression make it possible to fit large DNNs, such as AlexNet and VGGNet, fully in on-chip SRAM. But sparse networks compressed using existing encoding formats, like CSR or CSC, complex the computation at runtime due to their irregular memory access characteristics. In [1], we introduce a computation dataflow, stacked filters stationary dataflow (SFS), and a corresponding data encoding format, relative indexed compressed sparse filter format (CSF), to make the best of data sparsity, and simplify data handling at execution time. In this paper we present FPGA implementations of these methods. We implement several compact streaming fully connected (FC) and Convolutional (CONV) neural network processors to show their efficiency. Comparing with the state-of-the-art results [2,3,4], our methods achieve at least 2x improvement for computation efficiency per PE on most layers. Especially, our methods achieve 8x improvement on AlexNet layer CONV4 with 384 filters, and 11x improvement on VGG16 layer CONV5-3 with 512 filters.

cs.CV↗

arXiv2018

Towards Reliable (and Efficient) Job Executions in a Practical Geo-distributed Data Analytics System

Geo-distributed data analytics are increasingly common to derive useful information in large organisations. Naive extension of existing cluster-scale data analytics systems to the scale of geo-distributed data centers faces unique challenges including WAN bandwidth limits, regulatory constraints, changeable/unreliable runtime environment, and monetary costs. Our goal in this work is to develop a practical geo-distribued data analytics system that (1) employs an intelligent mechanism for jobs to efficiently utilize (adjust to) the resources (changeable environment) across data centers; (2) guarantees the reliability of jobs due to the possible failures; and (3) is generic and flexible enough to run a wide range of data analytics jobs without requiring any changes. To this end, we present a new, general geo-distributed data analytics system, HOUTU, that is composed of multiple autonomous systems, each operating in a sovereign data center. HOUTU maintains a job manager (JM) for a geo-distributed job in each data center, so that these replicated JMs could individually and cooperatively manage resources and assign tasks. Our experiments on the prototype of HOUTU running across four Alibaba Cloud regions show that HOUTU provides nearly efficient job performance as in the existing centralized architecture, and guarantees reliable job executions when facing failures.

cs.DC↗

arXiv2018

Stacked Filters Stationary Flow For Hardware-Oriented Acceleration Of Deep Convolutional Neural Networks

To address memory and computation resource limitations for hardware-oriented acceleration of deep convolutional neural networks (CNNs), we present a computation flow, stacked filters stationary flow (SFS), and a corresponding data encoding format, relative indexed compressed sparse filter format (CSF), to make the best of data sparsity, and simplify data handling at execution time. And we also propose a three dimensional Single Instruction Multiple Data (3D-SIMD) processor architecture to illustrate how to accelerate deep CNNs by taking advantage of SFS flow and CSF format. Comparing with the state-of-the-art result (Han et al., 2016b), our methods achieve 1.11x improvement in reducing the storage required by AlexNet, and 1.09x improvement in reducing the storage required by SqueezeNet, without loss of accuracy on the ImageNet dataset. Moreover, using these approaches, chip area for logics handling irregular sparse data access can be saved. Comparing with the 2D-SIMD processor structures in DVAS, ENVISION, etc., our methods achieve about 3.65x processing element (PE) array utilization rate improvement (from 26.4\% to 96.5\%) on the data from Deep Compression on AlexNet.

cs.CV↗

arXiv2018

Computation of Optimal Control Problems with Terminal Constraint via Modified Evolution Partial Differential Equation

The Variation Evolving Method (VEM), which seeks the optimal solutions with the variation evolution principle, is further developed to be more flexible in solving the Optimal Control Problems (OCPs) with terminal constraint. With the first-order stable dynamics to eliminate the infeasibilities, the Modified Evolution Partial Differential Equation (MEPDE) that is valid in the infeasible solution domain is proposed, and a Lyapunov functional is constructed to theoretically ensure its validity. In particular, it is proved that even with the infinite-time convergence dynamics, the violated terminal inequality constraints, which are inactive for the optimal solution, will enter the feasible domain in finite time. Through transforming the MEPDE to the finite-dimensional Initial-value Problem (IVP) with the semi-discrete method, the OCPs may be solved with common Ordinary Differential Equation (ODE) numerical integration methods. Illustrative examples are presented to show the effectiveness of the proposed method.

eess.SY↗

arXiv2018

On the Computation of Optimal Control Problems with Terminal Inequality Constraint via Variation Evolution

Studies regarding the computation of Optimal Control Problems (OCPs) with terminal inequality constraint, under the frame of the Variation Evolving Method (VEM), are carried out. The attributes of equality constraints and inequality constraints in the generalized optimization problem is traversed, and the intrinsic relations to the multipliers are uncovered. Upon these preliminaries, the right Evolution Partial Differential Equation (EPDE) is derived, and the costate-free optimality conditions are established. Besides the analytic expression for the costates in the classic treatment, they also reveal the analytic relations between the states, the controls and the (Lagrange and KKT) multipliers, which adjoin the terminal (equality and inequality) constraints. Moreover, in solving the transformed Initial-value Problems (IVPs) with common Ordinary Differential Equation (ODE) integration methods, the numerical soft barrier is proposed to eliminate the numerical error resulting from the suddenly triggered inequality constraint and it is shown to be effective.

eess.SY↗

arXiv2018

Giant photo-effect in proton transport through graphene membranes

Graphene has recently been shown to be permeable to thermal protons, the nuclei of hydrogen atoms, which sparked interest in its use as a proton-conducting membrane in relevant technologies. However, the influence of light on proton permeation remains unknown. Here we report that proton transport through Pt-nanoparticle-decorated graphene can be enhanced strongly by illuminating it with visible light. Using electrical measurements and mass spectrometry, we find a photoresponsivity of 10^4 A W-1, which translates into a gain of 10^4 protons per photon with response times in the microsecond range. These characteristics are competitive with those of state-of-the-art photodetectors that are based on electron transport using silicon and novel two-dimensional materials. The photo-proton effect can be important for graphene's envisaged use in fuel cells and hydrogen isotope separation. Our observations can also be of interest for other applications such as light-induced water splitting, photocatalysis and novel photodetectors.

physics.app-ph↗

arXiv2018

Random Euler Complex-Valued Nonlinear Filters

Over the last decade, both the neural network and kernel adaptive filter have successfully been used for nonlinear signal processing. However, they suffer from high computational cost caused by their complex/growing network structures. In this paper, we propose two random Euler filters for complex-valued nonlinear filtering problem, i.e., linear random Euler complex-valued filter (LRECF) and its widely-linear version (WLRECF), which possess a simple and fixed network structure. The transient and steady-state performances are studied in a non-stationary environment. The analytical minimum mean square error (MSE) and optimum step-size are derived. Finally, numerical simulations on complex-valued nonlinear system identification and nonlinear channel equalization are presented to show the effectiveness of the proposed methods.

stat.ML↗

arXiv2018

Computation of Optimal Control Problems with Terminal Constraint via Variation Evolution

Enlightened from the inverse consideration of the stable continuous-time dynamics evolution, the Variation Evolving Method (VEM) analogizes the optimal solution to the equilibrium point of an infinite-dimensional dynamic system and solves it in an asymptotically evolving way. In this paper, the compact version of the VEM is further developed for the computation of Optimal Control Problems (OCPs) with terminal constraint. The corresponding Evolution Partial Differential Equation (EPDE), which describes the variation motion towards the optimal solution, is derived, and the costate-free optimality conditions are established. The explicit analytic expressions of the costates and the Lagrange multipliers adjoining the terminal constraint, related to the states and the control variables, are presented. With the semi-discrete method in the field of PDE numerical calculation, the EPDE is discretized as finite-dimensional Initial-value Problems (IVPs) to be solved, with common Ordinary Differential Equation (ODE) numerical integration methods.

eess.SY↗

arXiv2017

Aircraft trajectory control with feedback linearization for general nonlinear system

The feedback linearization method is further developed for the controller design on general nonlinear systems. Through the Lyapunov stability theory, the intractable nonlinear implicit algebraic control equations are effectively solved, and the asymptotically tracking performance is guaranteed. Moreover, it is proved that the controller may be used in an inverse-free version to the set-point control. With this method, a nonlinear aircraft outer-loop trajectory controller is developed. For the concern regarding the controller's robustness, the integral control technique is combined to counteract the adverse effect from modeling errors. Simulation results verify the well performance of the proposed controller.

eess.SY↗

arXiv2017

Optimal Control Computation via Evolution Partial Differential Equation with Arbitrary Definite Conditions

The compact Variation Evolving Method (VEM) that originates from the continuous-time dynamics stability theory seeks the optimal solutions with variation evolution principle. It is further developed to be more flexible in solving the Optimal Control Problems (OCPs), by relaxing the definite conditions from a feasible solution to an arbitrary one for the derived Evolution Partial Differential Equation (EPDE). To guarantee the validity, an unconstrained Lyapunov functional that has the same minimum as the original OCP is constructed, and it ensures the evolution towards the optimal solution from infeasible solutions. With the semi-discrete method, the EPDE is transformed to the finite-dimensional Initial-value Problem (IVP), and then solved with common Ordinary Differential Equation (ODE) numerical integration methods. Illustrative examples are presented to show the effectiveness of the proposed method.

eess.SY↗

arXiv2017

Feature Enhancement Network: A Refined Scene Text Detector

In this paper, we propose a refined scene text detector with a \textit{novel} Feature Enhancement Network (FEN) for Region Proposal and Text Detection Refinement. Retrospectively, both region proposal with \textit{only} $3\times 3$ sliding-window feature and text detection refinement with \textit{single scale} high level feature are insufficient, especially for smaller scene text. Therefore, we design a new FEN network with \textit{task-specific}, \textit{low} and \textit{high} level semantic features fusion to improve the performance of text detection. Besides, since \textit{unitary} position-sensitive RoI pooling in general object detection is unreasonable for variable text regions, an \textit{adaptively weighted} position-sensitive RoI pooling layer is devised for further enhancing the detecting accuracy. To tackle the \textit{sample-imbalance} problem during the refinement stage, we also propose an effective \textit{positives mining} strategy for efficiently training our network. Experiments on ICDAR 2011 and 2013 robust text detection benchmarks demonstrate that our method can achieve state-of-the-art results, outperforming all reported methods in terms of F-measure.

cs.CV↗

arXiv2017

Computation of time-optimal control problem with variation evolution principle

An effective form of the Variation Evolving Method (VEM), which originates from the continuous-time dynamics stability theory, is developed for the classic time-optimal control problem with control constraint. Within the mathematic derivation, the Pontryagin's Minimum Principle (PMP) optimality conditions are used. Techniques including limited integrator and corner points are introduced to capture the right solution. The variation dynamic evolving equation may be reformulated as the Partial Differential Equation (PDE), and then discretized as finite-dimensional Initial-value Problem (IVP) to be solved with common Ordinary Differential Equation (ODE) integration methods. An illustrative example is solved to show the effectiveness of the method. In particular, the VEM is further developed to be more flexible in treating the boundary conditions of the Optimal Control Problem (OCP), by initializing the transformed IVP with arbitrary initial values of variables.

eess.SY↗

arXiv2017

Asymptotic properties of Banach spaces and coarse quotient maps

We give a quantitative result about asymptotic moduli of Banach spaces under coarse quotient maps. More precisely, we prove that if a Banach space $Y$ is a coarse quotient of a subset of a Banach space $X$, where the coarse quotient map is coarse Lipschitz, then the ($β$)-modulus of $X$ is bounded by the modulus of asymptotic uniform smoothness of $Y$ up to some constants. In particular, if the coarse quotient map is a coarse homeomorphism, then the modulus of asymptotic uniform convexity of $X$ is bounded by the modulus of asymptotic uniform smoothness of $Y$ up to some constants.

math.FA↗

arXiv2017

Dynamic backstepping control for pure-feedback nonlinear systems

A dynamic backstepping method is proposed to design controllers for nonlinear systems in the pure-feedback form, for which the traditional backstepping method suffers from solving the implicit nonlinear algebraic equation. The idea of this method is to augment the (virtual) controls as states during each recursive step. As new dynamics are included in the design, the resulting controller is in the dynamic feedback form. Procedure of deriving the controller is detailed, and one more Lyapunov design is executed in each step compared with the traditional backstepping method. Under appropriate assumptions, the proposed control scheme achieves the uniformly asymptotically stability. The effectiveness of this method is illustrated by the stabilization and tracking numerical examples.

eess.SY↗

arXiv2017

Deep Generalized Canonical Correlation Analysis

We present Deep Generalized Canonical Correlation Analysis (DGCCA) -- a method for learning nonlinear transformations of arbitrarily many views of data, such that the resulting transformations are maximally informative of each other. While methods for nonlinear two-view representation learning (Deep CCA, (Andrew et al., 2013)) and linear many-view representation learning (Generalized CCA (Horst, 1961)) exist, DGCCA is the first CCA-style multiview representation learning technique that combines the flexibility of nonlinear (deep) representation learning with the statistical power of incorporating information from many independent sources, or views. We present the DGCCA formulation as well as an efficient stochastic optimization algorithm for solving it. We learn DGCCA representations on two distinct datasets for three downstream tasks: phonetic transcription from acoustic and articulatory measurements, and recommending hashtags and friends on a dataset of Twitter users. We find that DGCCA representations soundly beat existing methods at phonetic transcription and hashtag recommendation, and in general perform no worse than standard linear many-view techniques.

cs.LG↗

arXiv2017

Robust Tracking Guidance for Zero Propellant Maneuver

The Zero Propellant Maneuver (ZPM) maneuvers the space station by large angle, utilizing the Control Momentum Gyroscopes (CMGs) only. A robust tracking guidance strategy is proposed to enhance its performance. It is distinguished from the traditional trajectory tracking guidance in that the reference trajectory is adjusted on-line, under the inspiration of eliminating the discrepancy on the total angular momentum of the space station system. The Lyapunov controller is developed to adjust the attitude trajectory and further redesigned for a better performance based on an interesting physical phenomenon, which is taken advantage of by coupling the components of state vector. The adjusted trajectory is then tracked to reach the target states of maneuver. Simulations results show that the disturbance effects arising from initial state errors, parameter uncertainty and modeling errors are attenuated or even eliminated, which verifies the robustness of the guidance strategy.

eess.SY↗

arXiv2017

Ordinal Common-sense Inference

Humans have the capacity to draw common-sense inferences from natural language: various things that are likely but not certain to hold based on established discourse, and are rarely stated explicitly. We propose an evaluation of automated common-sense inference based on an extension of recognizing textual entailment: predicting ordinal human responses on the subjective likelihood of an inference holding in a given context. We describe a framework for extracting common-sense knowledge from corpora, which is then used to construct a dataset for this ordinal entailment task. We train a neural sequence-to-sequence model on this dataset, which we use to score and generate possible inferences. Further, we annotate subsets of previously established datasets via our ordinal annotation protocol in order to then analyze the distinctions between these and what we have constructed.

cs.CL↗