arXiv subjects

Jun Yang

Publications and source records attributed to Jun Yang.

At least 397 records · Page 22Linked to original sources

arXiv2019

Fast-rate PAC-Bayes Generalization Bounds via Shifted Rademacher Processes

The developments of Rademacher complexity and PAC-Bayesian theory have been largely independent. One exception is the PAC-Bayes theorem of Kakade, Sridharan, and Tewari (2008), which is established via Rademacher complexity theory by viewing Gibbs classifiers as linear operators. The goal of this paper is to extend this bridge between Rademacher complexity and state-of-the-art PAC-Bayesian theory. We first demonstrate that one can match the fast rate of Catoni's PAC-Bayes bounds (Catoni, 2007) using shifted Rademacher processes (Wegkamp, 2003; Lecué and Mitchell, 2012; Zhivotovskiy and Hanneke, 2018). We then derive a new fast-rate PAC-Bayes bound in terms of the "flatness" of the empirical risk surface on which the posterior concentrates. Our analysis establishes a new framework for deriving fast-rate PAC-Bayes bounds and yields new insights on PAC-Bayesian theory.

cs.LG↗

arXiv2019

Performance Comparison of UCA and UCCA based Real-time Sound Source Localization Systems using Circular Harmonics SRP Method

Many sound source localization (SSL) algorithms based on circular microphone array (CMA), including uniform circular array (UCA) and uniform concentric circular array (UCCA), have been well developed and verified via computer simulations and offline processing. On the other hand, beamforming in the harmonic domain has been shown to be a very efficient tool for broadband sound source localization due to its frequency-invariant properties. In this paper, design, implementation, and performance of a real-time sound source localization system are discussed. Specifically, we analyze the effect of parameter settings and compare the performance between UCA and UCCA through real-time experiments in a real room. The proposed method shows significant improvement by using UCCA instead of UCA.

eess.AS↗

arXiv2019

FusionStitching: Boosting Execution Efficiency of Memory Intensive Computations for DL Workloads

Performance optimization is the art of continuous seeking a harmonious mapping between the application domain and hardware. Recent years have witnessed a surge of deep learning (DL) applications in industry. Conventional wisdom for optimizing such workloads mainly focus on compute intensive ops (GEMM, Convolution, etc). Yet we show in this work, that the performance of memory intensive computations is vital to E2E performance in practical DL models. We propose \emph{FusionStitching}, a optimization framework capable of fusing memory intensive \emph{elementwise}, \emph{reduction} and fine grained \emph{GEMM/Batched-GEMM} ops, with or without data dependences, into large computation units, then mapping and transforming them into efficient GPU kernels. We formulate the fusion plan optimization as an integer linear programming (ILP) problem, and propose a set of empirical heuristics to reduce the combinatorial search space. In order to map optimized fusion plans to hardware, we propose a technique to effectively compose various groups of computations into a single GPU kernel, by fully leveraging on chip resources like scratchpads or registers. Experimental results on six benchmarks and four industry scale practical models are encouraging. Overall, \emph{FusionStitching} can reach up to 5.7x speedup compared to Tensorflow baseline, and achieves 1.25x to 1.85x performance speedups compared to current state of the art, with 1.4x on average (geometric mean).

cs.DC↗

arXiv2019

Study on Asymmetric Diffraction of Acoustic Parity-Time-Symmetric Gratings Using Rigorous Coupled-Wave Analysis

In PT-symmetric gratings, asymmetric diffraction can be generated by modulating the ratio between imaginary and real refractive indices. In this paper, a rigorous coupled-wave analysis (RCWA) has been developed to analyze the diffraction properties of acoustic PT-symmetric gratings with two kinds of modulating approaches, including modulating the effective modulus and the effective density. Asymmetric diffraction with both Bragg incident angles and perpendicular incident angles is discussed by using the RCWA method. Results show that the modulation ratio for the diffraction vanishing point changes with the modulation amplitude differently for two kinds of modulating approaches. Moreover, the sound energy will be weaken or be enhanced at Bragg incident angles depending on the sign of incident angles and the modulation ratio.

physics.class-ph↗

arXiv2019

Sound Localization and Separation in Three-dimensional Space Using a Single Microphone with a Metamaterial Enclosure

Conventional approaches to sound localization and separation are based on microphone arrays in artificial systems. Inspired by the selective perception of human auditory system, we design a multi-source listening system which can separate simultaneous overlapping sounds and localize the sound sources in three-dimensional space, using only a single microphone with a metamaterial enclosure. The enclosure modifies the frequency response of the microphone in a direction-dependent way by giving each direction a signature. Thus, the information about the location and audio content of sound sources can be experimentally reconstructed from the modulated mixed signals using compressive sensing algorithm. Owing to the low computational complexity of the proposed reconstruction algorithm, the designed system can also be applied in source identification and tracking. The effectiveness of the system in multiple real scenarios has been proved through multiple random listening tests. The proposed metamaterial-based single-sensor listening system opens a new way of sound localization and separation, which can be applied to intelligent scene monitoring and robot audition.

cs.SD↗

arXiv2019

Sparking mashups to form multifunctional alloy nanoparticles

Synthesizing unconventional alloys remains challenging owing to seamless interactions between kinetics and thermodynamics. High entropy alloys (HEAs), for example, draw a fundamentally new concept to enable exploring unknown regions in phase diagrams. The exploration, however, is hindered by traditional metallurgies based on liquid-solid transformation. Vapor-solid transformation that is permissible on pressure-temperature phase diagrams, offers the most kinetically efficient pathway to form any desired alloy (e.g., HEA). Here, we report that a technique called "sparking mashups", which involves a rapidly quenched vapor source and induces unrestricted mixing for alloying 55 distinct types of ultrasmall nanoparticles (NPs) with controllable compositions. Unlike the precursor feed in wet chemistry, a microseconds-long oscillatory spark controls the vapour composition, which is eventually retained in the alloy NPs. The resulting NPs range from binary to HEAs with marked thermal stability at room temperature. We show that a nanosize-effect ensures such thermal stability and mimics the role of mixing entropy in HEAs. This discovery contradicts the traditional "smaller is less stable" view while enabling the elemental combinations that have never been alloyed to date. We even break the miscibility limits by mixing bulk-immiscible systems in alloy NPs. As powerful examples, we demonstrate the alloy NPs as both high-performance fuel-cell catalysts and building blocks for three-dimensional (3D) nanoprinting to construct HEA nanostructure arrays of various architectures and compositions. Our results form the basis of new rules for guiding HEA-NP synthesis and advancing catalysis and 3D printing to new frontiers.

physics.app-ph↗

arXiv2019

How do Planetary Radius and Gravity Influence the Surface Climate of Earth-like Planets?

About 4000 exoplanets have been confirmed since the year of 1992, and for most of the planets, the main parameters that can be measured are planetary radius and mass. Based on these two parameters, surface gravity can be estimated. In this work, we investigate the effects of varying radius and gravity on the climate of rapidly rotating terrestrial planets with assuming they have an ocean and Earth-like atmospheres (N$_2$, CO$_2$, and H$_2$O). Using a three-dimensional (3D) atmospheric general circulation model (GCM), we find that varying radius mainly influences the equator-to-pole surface temperature difference while varying gravity mainly influences the mean surface temperature. For planets of larger radii, the meridional atmospheric energy transport is weaker, which warms the tropics but cools high latitudes. For planets of larger gravities, the surface is globally cooler due to the fact that saturated vapor pressure depends on air temperature only and column water vapor mass is approximately equal to saturated vapor pressure over gravity multiplied by relative humidity. The relative humidity does not change much in our experiments. Ice albedo and water vapor feedbacks act to further amplify the effects of varying radius and gravity. These results suggest that radius and gravity are important factors for planetary climate, although they are not as critical as stellar flux, atmospheric composition and mass, and rotation rate. Future simulations are required to know how large the radius and gravity could influence the width of the habitable zone using 3D GCMs.

astro-ph.EP↗

arXiv2019

Characterizing Deep Learning Training Workloads on Alibaba-PAI

Modern deep learning models have been exploited in various domains, including computer vision (CV), natural language processing (NLP), search and recommendation. In practical AI clusters, workloads training these models are run using software frameworks such as TensorFlow, Caffe, PyTorch and CNTK. One critical issue for efficiently operating practical AI clouds, is to characterize the computing and data transfer demands of these workloads, and more importantly, the training performance given the underlying software framework and hardware configurations. In this paper, we characterize deep learning training workloads from Platform of Artificial Intelligence (PAI) in Alibaba. We establish an analytical framework to investigate detailed execution time breakdown of various workloads using different training architectures, to identify performance bottleneck. Results show that weight/gradient communication during training takes almost 62% of the total execution time among all our workloads on average. The computation part, involving both GPU computing and memory access, are not the biggest bottleneck based on collective behavior of the workloads. We further evaluate attainable performance of the workloads on various potential software/hardware mappings, and explore implications on software architecture selection and hardware configurations. We identify that 60% of PS/Worker workloads can be potentially sped up when ported to the AllReduce architecture exploiting the high-speed NVLink for GPU interconnect, and on average 1.7X speedup can be achieved when Ethernet bandwidth is upgraded from 25 Gbps to 100 Gbps.

cs.PF↗

arXiv2019

Selecting Data to Clean for Fact Checking: Minimizing Uncertainty vs. Maximizing Surprise

We study the optimization problem of selecting numerical quantities to clean in order to fact-check claims based on such data. Oftentimes, such claims are technically correct, but they can still mislead for two reasons. First, data may contain uncertainty and errors. Second, data can be "fished" to advance particular positions. In practice, fact-checkers cannot afford to clean all data and must choose to clean what "matters the most" to checking a claim. We explore alternative definitions of what "matters the most": one is to ascertain claim qualities (by minimizing uncertainty in these measures), while an alternative is just to counter the claim (by maximizing the probability of finding a counterargument). We show whether the two objectives align with each other, with important implications on when fact-checkers should exercise care in selective data cleaning, to avoid potential bias introduced by their desire to counter claims. We develop efficient algorithms for solving the various variants of the optimization problem, showing significant improvements over naive solutions. The problem is particularly challenging because the objectives in the fact-checking context are complex, non-linear functions over data. We obtain results that generalize to a large class of functions, with potential applications beyond fact-checking.

cs.DB↗

arXiv2019

Efficient Knowledge Graph Accuracy Evaluation

Estimation of the accuracy of a large-scale knowledge graph (KG) often requires humans to annotate samples from the graph. How to obtain statistically meaningful estimates for accuracy evaluation while keeping human annotation costs low is a problem critical to the development cycle of a KG and its practical applications. Surprisingly, this challenging problem has largely been ignored in prior research. To address the problem, this paper proposes an efficient sampling and evaluation framework, which aims to provide quality accuracy evaluation with strong statistical guarantee while minimizing human efforts. Motivated by the properties of the annotation cost function observed in practice, we propose the use of cluster sampling to reduce the overall cost. We further apply weighted and two-stage sampling as well as stratification for better sampling designs. We also extend our framework to enable efficient incremental evaluation on evolving KG, introducing two solutions based on stratified sampling and a weighted variant of reservoir sampling. Extensive experiments on real-world datasets demonstrate the effectiveness and efficiency of our proposed solution. Compared to baseline approaches, our best solutions can provide up to 60% cost reduction on static KG evaluation and up to 80% cost reduction on evolving KG evaluation, without loss of evaluation quality.

cs.DB↗

arXiv2019

X-ray differential phase contrast imaging on asymmetric dual-phase grating interferometer with source grating: theory and experiment

Recently, the dual-phase grating based X-ray differential phase contrast imaging technique has shown better radiation dose efficiency performance than the Talbot-Lau system. In this paper, we provide a theoretical analyses framework derived from wave optics to ease the design of such interferometer systems, including the inter-grating distances, the diffraction fringe period, the phase grating periods, and especially the source grating period if a medical grade X-ray tube with large focal spot is utilized. In addition, a geometrical explanation of the dual-phase grating system similar to the standard thin lens imaging theory is derived with an optical symmetry assumption for the first time. Finally, both numerical and experimental studies have been performed to validate the theory.

physics.optics↗

arXiv2019

A Sustainable Multi-modal Multi-layer Emotion-aware Service at the Edge

Limited by the computational capabilities and battery energy of terminal devices and network bandwidth, emotion recognition tasks fail to achieve good interactive experience for users. The intolerable latency for users also seriously restricts the popularization of emotion recognition applications in the edge environments such as fatigue detection in auto-driving. The development of edge computing provides a more sustainable solution for this problem. Based on edge computing, this article proposes a multi-modal multi-layer emotion-aware service (MULTI-EASE) architecture that considers user's facial expression and voice as a multi-modal data source of emotion recognition, and employs the intelligent terminal, edge server and cloud as multi-layer execution environment. By analyzing the average delay of each task and the average energy consumption at the mobile device, we formulate a delay-constrained energy minimization problem and perform a task scheduling policy between multiple layers to reduce the end-to-end delay and energy consumption by using an edge-based approach, further to improve the users' emotion interactive experience and achieve energy saving in edge computing. Finally, a prototype system is also implemented to validate the architecture of MULTI-EASE, the experimental results show that MULTI-EASE is a sustainable and efficient platform for emotion analysis applications, and also provide a valuable reference for dynamic task scheduling under MULTI-EASE architecture.

cs.NI↗

arXiv2019

On the non-degeneracy of radial vortex solutions for a coupled Ginzburg-Landau system

For the following Ginzburg-Landau system in ${\mathbb R}^2$ \begin{align*} \begin{cases} -Δw^+ +\Big[A_+\big(|w^+|^2-{t^+}^2\big)+B\big(|w^-|^2-{t^-}^2\big)\Big]w^+=0, \\[3mm] -Δw^- +\Big[A_-\big(|w^-|^2-{t^-}^2\big)+B\big(|w^+|^2-{t^+}^2\big)\Big]w^-=0, \end{cases} \end{align*} with constraints $ A_+, A_->0$, $B<0$, $B^2 0$, we will concern its linearized operator ${\mathcal L}$ around the radially symmetric solution $w(x)=(w^+, w^-): {\mathbb R}^2 \rightarrow\mathbb{C}^2$ of degree pair $(1, 1)$ and prove the non-degeneracy result: the kernel of ${\mathcal L}$ is spanned by $\big\{\frac{\partial w}{\partial{x_1}}, \frac{\partial w}{\partial{x_2}}\big\}$ in a natural Hilbert space. As an application of the non-degeneracy result, a solvability theory for the linearized operator ${\mathcal L}$ will be given.

math.AP↗

arXiv2019

Experimental demonstration of asymmetric diffraction based on a passive parity-time-symmetric acoustic grating

Passive parity-time-symmetric medium provides a feasible scheme to investigate non-Hermitian systems experimentally. Here, we design a passive PT-symmetric acoustic grating with a period equal to exact PT-symmetric medium. This treatment enhances the diffraction ability of a passive PT-symmetric grating with more compact modulation. Above all, it eliminates the first-order disturbance of previous design in diffraction grating. Additional cavities and small leaked holes on top plate in a 2D waveguide are used to construct a parity-time-symmetric potential. The combining between additional cavities and leaked holes makes it possible to modulate the real and imaginary parts of refractive index simultaneously. When the real and imaginary parts of refractive index are balanced in modulation, asymmetric diffraction can be observed between a pair of oblique incident waves. This demonstration provides a feasible way to construct passive parity-time-symmetric acoustic medium. It opens new possibilities for further investigation of acoustic wave control in non-Hermitian systems.

physics.class-ph↗

arXiv2019

Solver for the electronic V-representation problem of time-dependent density functional theory

One route to numerically propagating quantum systems is time-dependent density functional theory (TDDFT). The application of TDDFT to a particular system's time evolution is predicated on $V$-representability which we have analyzed in a previous publication. Here we describe a newly developed solver for the scalar time-dependent Kohn-Sham potential. We present and interpret the force-balance equation central to our numerical method, describe details of its implementation, and present illustrative numerical results for one- and two-electron systems. A new characterization of $V$-representability for one-electron systems is also included along with possible improvements and future directions.

quant-ph↗

arXiv2019

Explaining Wrong Queries Using Small Examples

For testing the correctness of SQL queries, e.g., evaluating student submissions in a database course, a standard practice is to execute the query in question on some test database instance and compare its result with that of the correct query. Given two queries $Q_1$ and $Q_2$, we say that a database instance $D$ is a counterexample (for $Q_1$ and $Q_2$) if $Q_1(D)$ differs from $Q_2(D)$; such a counterexample can serve as an explanation of why $Q_1$ and $Q_2$ are not equivalent. While the test database instance may serve as a counterexample, it may be too large or complex to read and understand where the inequivalence comes from. Therefore, in this paper, given a known counterexample $D$ for $Q_1$ and $Q_2$, we aim to find the smallest counterexample $D' \subseteq D$ where $Q_1(D') \neq Q_2(D')$. The problem in general is NP-hard. We give a suite of algorithms for finding the smallest counterexample for different classes of queries, some more tractable than others. We also present an efficient provenance-based algorithm for SPJUD queries that uses a constraint solver, and extend it to more complex queries with aggregation, group-by, and nested queries. We perform extensive experiments indicating the effectiveness and scalability of our solution on student queries from an undergraduate database course and on queries from the TPC-H benchmark. We also report a user study from the course where we deployed our tool to help students with an assignment on relational algebra.

cs.DB↗

arXiv2019

Ready Player One: UAV Clustering based Multi-Task Offloading for Vehicular VR/AR Gaming

With rapid development of unmanned aerial vehicle (UAV) technology, application of the UAVs for task offloading has received increasing interest in the academia. However, real-time interaction between one UAV and the mobile edge computing (MEC) node is required for processing the tasks of mobile end users, which significantly increases the system overhead and is unable to meet the demands of large-scale artificial intelligence (AI) based applications. To tackle this problem, in this article, we propose a new architecture for UAV clustering to enable efficient multi-modal multi-task task offloading. By the proposed architecture, the computing, caching and communication resources are collaboratively optimized using AI based decision-making. This not only increases the efficiency of UAV clusters, but also provides insight into the fusion of computation and communication.

cs.NI↗

arXiv2019

Reconfigurable Multitask Audio Dynamics Processing Scheme

Automatic speech recognition (ASR), audio quality, and loudness are key performance indicators (KPIs) in smart speakers. To improve all these KPIs, audio dynamics processing is a crucial component in related systems. Unfortunately, single-band and existing multiband dynamics processing (MBDP) schemes fail to maximize bass and loudness but even produce unwanted peaks, distortions, and nonlinear echo so that an optimized ASR performance cannot be achieved. It has been a goal in both industry and academia to find a better audio dynamics processing for mitigating these problems. To provide such a desired solution, this paper proposes a novel reconfigurable multitask MBDP scheme through a global optimization framework. Through extensive testing, we show the accuracy and effectiveness of the proposed scheme in terms of bass and loudness maximization, distortion and nonlinear echo reduction, browning-out prevention, and significant ASR performance improvement.

eess.SP↗