arXiv subjects

Jia Liu

Publications and source records attributed to Jia Liu.

At least 379 records · Page 21Linked to original sources

arXiv2019

Latent Class Model with Application to Speaker Diarization

In this paper, we apply a latent class model (LCM) to the task of speaker diarization. LCM is similar to Patrick Kenny's variational Bayes (VB) method in that it uses soft information and avoids premature hard decisions in its iterations. In contrast to the VB method, which is based on a generative model, LCM provides a framework allowing both generative and discriminative models. The discriminative property is realized through the use of i-vector (Ivec), probabilistic linear discriminative analysis (PLDA), and a support vector machine (SVM) in this work. Systems denoted as LCM-Ivec-PLDA, LCM-Ivec-SVM, and LCM-Ivec-Hybrid are introduced. In addition, three further improvements are applied to enhance its performance. 1) Adding neighbor windows to extract more speaker information for each short segment. 2) Using a hidden Markov model to avoid frequent speaker change points. 3) Using an agglomerative hierarchical cluster to do initialization and present hard and soft priors, in order to overcome the problem of initial sensitivity. Experiments on the National Institute of Standards and Technology Rich Transcription 2009 speaker diarization database, under the condition of a single distant microphone, show that the diarization error rate (DER) of the proposed methods has substantial relative improvements compared with mainstream systems. Compared to the VB method, the relative improvements of LCM-Ivec-PLDA, LCM-Ivec-SVM, and LCM-Ivec-Hybrid systems are 23.5%, 27.1%, and 43.0%, respectively. Experiments on our collected database, CALLHOME97, CALLHOME00 and SRE08 short2-summed trial conditions also show that the proposed LCM-Ivec-Hybrid system has the best overall performance.

eess.AS↗

arXiv2019

Long-lived particles at the LHC: catching them in time

We explore the physics potential of using precision timing information at the LHC in searches for long-lived particles (LLPs). In comparison with the light Standard Model particles, the decay products of massive LLPs arrive at detectors with time delays around the nanosecond scale. We propose new strategies to take advantage of this time delay feature by using initial state radiation to timestamp the collision event and require at least one LLP to decay within the detector. This search strategy is effective for a broad range of models. In addition to outlining this general approach, we demonstrate its effectiveness with the projected reach for two benchmark scenarios: Higgs decaying into a pair of LLPs, and pair production of long-lived neutralinos in the gauge mediated supersymmetry breaking models. Our strategy increases the sensitivity to the lifetime of the LLP by two orders of magnitude or more and particularly exhibits a better behavior with a linear dependence on lifetime in the large lifetime region compared to traditional LLP searches. The timing information significantly reduces the Standard Model background and provides a powerful new dimension for LLP searches.

hep-ph↗

arXiv2019

A $\pmbν$ Solution to the Strong CP Problem

We present a new solution to the strong CP problem in which the imaginary component of the up quark mass, $\mathcal{I}[m_u]$, acquires a tiny, but non-vanishing value. This is achieved via a Dirac seesaw mechanism, which is also responsible for the generation of the small neutrino masses. Consistency with the observed value of the up quark mass is achieved via instanton contributions arising from QCD-like interactions. In this framework, the value of the neutron electric dipole moment is directly related to $\mathcal{I}[m_u]$, which, due to its common origin with the neutrino masses, implies that the neutron electric dipole moment is likely to be measured in the next round of experiments. We also present a supersymmetric extension of this Dirac seesaw model to stabilize the hierarchy among the scalar mass scales involved in this new mechanism.

hep-ph↗

arXiv2019

Large Margin Softmax Loss for Speaker Verification

In neural network based speaker verification, speaker embedding is expected to be discriminative between speakers while the intra-speaker distance should remain small. A variety of loss functions have been proposed to achieve this goal. In this paper, we investigate the large margin softmax loss with different configurations in speaker verification. Ring loss and minimum hyperspherical energy criterion are introduced to further improve the performance. Results on VoxCeleb show that our best system outperforms the baseline approach by 15\% in EER, and by 13\%, 33\% in minDCF08 and minDCF10, respectively.

cs.SD↗

arXiv2019

A light complex scalar for the electron and muon anomalous magnetic moments

The anomalous magnetic moments of the electron and the muon are interesting observables, since they can be measured with great precision and their values can be computed with excellent accuracy within the Standard Model (SM). The current experimental measurement of this quantities show a deviation of a few standard deviations with respect to the SM prediction, which may be a hint of new physics. The fact that the electron and the muon masses differ by two orders of magnitude and the deviations have opposite signs makes it difficult to find a common origin of these anomalies. In this work we introduce a complex singlet scalar charged under a Peccei-Quinn-like (PQ) global symmetry together with the electron transforming chirally under the same symmetry. In this realization, the CP-odd scalar couples to electron only, while the CP-even part can couple to muons and electrons simultaneously. In addition, the CP-odd scalar can naturally be much lighter than the CP-even scalar, as a pseudo-Goldstone boson of the PQ-like symmetry, leading to an explanation of the suppression of the electron anomalous magnetic moment with respect to the SM prediction due to the CP-odd Higgs effect dominance, as well as an enhancement of the muon one induced by the CP-even component.

hep-ph↗

arXiv2019

Can We Achieve Fresh Information with Selfish Users in Mobile Crowd-Learning?

The proliferation of smart mobile devices has spurred an explosive growth of mobile crowd-learning services, where service providers rely on the user community to voluntarily collect, report, and share real-time information for a collection of scattered points of interest. A critical factor affecting the future large-scale adoption of such mobile crowd-learning applications is the freshness of the crowd-learned information, which can be measured by a metric termed ``age-of-information'' (AoI). However, we show that the AoI of mobile crowd-learning could be arbitrarily bad under selfish users' behaviors if the system is poorly designed. This motivates us to design efficient reward mechanisms to incentivize mobile users to report information in time, with the goal of keeping the AoI and congestion level of each PoI low. Toward this end, we consider a simple linear AoI-based reward mechanism and analyze its AoI and congestion performances in terms of price of anarchy (PoA), which characterizes the degradation of the system efficiency due to selfish behavior of users. Remarkably, we show that the proposed mechanism achieves the optimal AoI performance asymptotically in a deterministic scenario. Further, we prove that the proposed mechanism achieves a bounded PoA in general stochastic cases, and the bound only depends on system parameters. Particularly, when the service rates of PoIs are symmetric in stochastic cases, the achieved PoA is upper-bounded by $1/2$ asymptotically. Collectively, this work advances our understanding of information freshness in mobile crowd-learning systems.

cs.SI↗

arXiv2018

Constraining neutrino mass with weak lensing Minkowski Functionals

The presence of massive neutrinos affects structure formation, leaving imprints on large-scale structure observables such as the weak lensing field. The common lensing analyses with two-point statistics are insensitive to the large amount of non-Gaussian information in the density field. We investigate non-Gaussian tools, in particular the Minkowski Functionals (MFs)---morphological descriptors including area, perimeter, and genus---in an attempt to recover the higher-order information. We use convergence maps from the Cosmological Massive Neutrino Simulations (\texttt{MassiveNus}) and assume galaxy noise, density, and redshift distribution for an LSST-like survey. We show that MFs are sensitive to the neutrino mass sum, and the sensitivity is redshift dependent and is non-Gaussian. We find that redshift tomography significantly improves the constraints on neutrino mass for MFs, compared to the improvements for the power spectrum. We attribute this to the stronger redshift dependence of neutrino effects on small scales. We then build an emulator to model the power spectrum and MFs, and study the constraints on $[M_ν$, $Ω_{m}$, $A_{s}]$ from the power spectrum, MFs, and their combination. We show that MFs significantly outperform the power spectrum in constraining neutrino mass, by more than a factor of four. However, a thorough study of the impact from systematics such as baryon physics and galaxy shape and redshift biases will be important to realize the full potential of MFs.

astro-ph.CO↗

arXiv2018

Localized polarons and conductive charge carriers: understanding CaCu$_{3}$Ti$_{4}$O$_{12}$ over a broad temperature range

CaCu$_{3}$Ti$_{4}$O$_{12}$ (CCTO) has a large dielectric permittivity that is independent of the probing frequency near the room temperature, which complicated due to the existence of several dynamic processes. Here, we consider the combined effects of localized charge carriers (polarons) and thermally activated charge carriers using a recently proposed statistical model to fit and understand the permittivity of CCTO measured at different frequencies over the whole temperature range accessible by our experiments. We found that the small permittivity at the lowest temperature is related to polaron frozen, while at higher temperatures the rapid increase is associated with the thermal excitation of polarons inducing the Maxwell-Wagner effect, and the final increase of the permittivity is attributed to the thermally activated conductivity. Such analysis enables us to separate the contributions from localized polarons and conductive charge carriers and quantify their activation energies.

cond-mat.mtrl-sci↗

arXiv2018

Multiobjective Optimization Training of PLDA for Speaker Verification

Most current state-of-the-art text-independent speaker verification systems take probabilistic linear discriminant analysis (PLDA) as their backend classifiers. The parameters of PLDA are often estimated by maximizing the objective function, which focuses on increasing the value of log-likelihood function, but ignoring the distinction between speakers. In order to better distinguish speakers, we propose a multi-objective optimization training for PLDA. Experiment results show that the proposed method has more than 10% relative performance improvement in both EER and MinDCF on the NIST SRE14 i-vector challenge dataset, and about 20% relative performance improvement in EER on the MCE18 dataset.

cs.SD↗

arXiv2018

Steganography Security: Principle and Practice

This paper focuses on several theoretical issues and principles in steganography security, and defines four security levels by analyzing the corresponding algorithm instances. In the theoretical analysis, we discuss the differences between steganography security and watermarking security. The two necessary conditions for the steganography security are obtained. Under the current technology situation, we then analyze the indistinguishability of the cover and stego-cover, and consider that the steganography security should rely on the key secrecy with algorithms open. By specifying the role of key in steganography, the necessary conditions for a secure steganography algorithm in theory are formally presented. When analyzing the security instances, we have classified the steganalysis attacks according to their variable access to the steganography system, and then defined the four security levels. The higher level security one has, the higher level attacks one can resist. We have also presented algorithm instances based on current technical conditions, and analyzed their data hiding process, security level, and practice requirements.

cs.MM↗

arXiv2018

Constraining Neutrino Mass with the Tomographic Weak Lensing Bispectrum

We explore the effect of massive neutrinos on the weak lensing shear bispectrum using the Cosmological Massive Neutrino Simulations. We find that the primary effect of massive neutrinos is to suppress the amplitude of the bispectrum with limited effect on the bispectrum shape. The suppression of the bispectrum amplitude is a factor of two greater than the suppression of the small scale power-spectrum. For an LSST-like weak lensing survey that observes half of the sky with five tomographic redshift bins, we explore the constraining power of the bispectrum on three cosmological parameters: the sum of the neutrino mass $\sum m_ν$, the matter density $Ω_m$ and the amplitude of primordial fluctuations $A_s$. Bispectrum measurements alone provide similar constraints to the power spectrum measurements and combining the two probes leads to significant improvements than using the latter alone. We find that the joint constraints tighten the power spectrum $95\%$ constraints by $\sim 32\%$ for $\sum m_ν$, $13\%$ for $Ω_m$ and $57\%$ for $A_s$ .

astro-ph.CO↗

arXiv2018

Constraining neutrino mass with tomographic weak lensing peak counts

Massive cosmic neutrinos change the structure formation history by suppressing perturbations on small scales. Weak lensing data from galaxy surveys probe the structure evolution and thereby can be used to constrain the total mass of the three active neutrinos. However, much of the information is at small scales where the dynamics are nonlinear. Traditional approaches with second order statistics thus fail to fully extract the information in the lensing field. In this paper, we study constraints on the neutrino mass sum using lensing peak counts, a statistic which captures non-Gaussian information beyond the second order. We use the ray-traced weak lensing mocks from the Cosmological Massive Neutrino Simulations (MassiveNuS), and apply LSST-like noise. We discuss the effects of redshift tomography, multipole cutoff $\ell_{\rm max}$ for the power spectrum, smoothing scale for the peak counts, and constraints from peaks of different heights. We find that combining peak counts with the traditional lensing power spectrum can improve the constraint on neutrino mass sum, $Ω_m$, and $A_s$ by 39%, 32%, and 60%, respectively, over that from the power spectrum alone.

astro-ph.CO↗

arXiv2018

Constraining neutrino mass with tomographic weak lensing one-point probability distribution function and power spectrum

We study the constraints on neutrino mass sum (M_nu) from the one-point probability distribution function (PDF) and power spectrum of weak lensing measurements for an LSST-like survey, using the MassiveNuS simulations. The PDF provides access to non-Gaussian information beyond the power spectrum. It is particularly sensitive to nonlinear growth on small scales, where massive neutrinos also have the largest effect. We find that tomography helps improve the constraint on M_nu by 14% and 32% for the power spectrum and the PDF, respectively, compared to a single redshift bin. The PDF alone outperforms the power spectrum in constraining M_nu. When the two statistics are combined, the constraint is further tightened by 35%. We conclude that weak lensing PDF is complementary to the power spectrum and has the potential to become a powerful tool for constraining neutrino mass.

astro-ph.CO↗

arXiv2018

Poisoning Attacks to Graph-Based Recommender Systems

Recommender system is an important component of many web services to help users locate items that match their interests. Several studies showed that recommender systems are vulnerable to poisoning attacks, in which an attacker injects fake data to a given system such that the system makes recommendations as the attacker desires. However, these poisoning attacks are either agnostic to recommendation algorithms or optimized to recommender systems that are not graph-based. Like association-rule-based and matrix-factorization-based recommender systems, graph-based recommender system is also deployed in practice, e.g., eBay, Huawei App Store. However, how to design optimized poisoning attacks for graph-based recommender systems is still an open problem. In this work, we perform a systematic study on poisoning attacks to graph-based recommender systems. Due to limited resources and to avoid detection, we assume the number of fake users that can be injected into the system is bounded. The key challenge is how to assign rating scores to the fake users such that the target item is recommended to as many normal users as possible. To address the challenge, we formulate the poisoning attacks as an optimization problem, solving which determines the rating scores for the fake users. We also propose techniques to solve the optimization problem. We evaluate our attacks and compare them with existing attacks under white-box (recommendation algorithm and its parameters are known), gray-box (recommendation algorithm is known but its parameters are unknown), and black-box (recommendation algorithm is unknown) settings using two real-world datasets. Our results show that our attack is effective and outperforms existing attacks for graph-based recommender systems. For instance, when 1% fake users are injected, our attack can make a target item recommended to 580 times more normal users in certain scenarios.

cs.IR↗

arXiv2018

PAC Ranking from Pairwise and Listwise Queries: Lower Bounds and Upper Bounds

This paper explores the adaptive (active) PAC (probably approximately correct) top-$k$ ranking (i.e., top-$k$ item selection) and total ranking problems from $l$-wise ($l\geq 2$) comparisons under the multinomial logit (MNL) model. By adaptively choosing sets to query and observing the noisy output of the most favored item of each query, we want to design ranking algorithms that recover the top-$k$ or total ranking using as few queries as possible. For the PAC top-$k$ ranking problem, we derive a lower bound on the sample complexity (aka number of queries), and propose an algorithm that is sample-complexity-optimal up to an $O(\log(k+l)/\log{k})$ factor. When $l=2$ (i.e., pairwise comparisons) or $l=O(poly(k))$, this algorithm matches the lower bound. For the PAC total ranking problem, we derive a tight lower bound, and propose an algorithm that matches the lower bound. When $l=2$, the MNL model reduces to the popular Plackett-Luce (PL) model. In this setting, our results still outperform the state-of-the-art both theoretically and numerically. We also compare our algorithms with the state-of-the-art using synthetic data as well as real-world data to verify the efficiency of our algorithms.

cs.LG↗

arXiv2018

Investigation of Frame Alignments for GMM-based Digit-prompted Speaker Verification

Frame alignments can be computed by different methods in GMM-based speaker verification. By incorporating a phonetic Gaussian mixture model (PGMM), we are able to compare the performance using alignments extracted from the deep neural networks (DNN) and the conventional hidden Markov model (HMM) in digit-prompted speaker verification. Based on the different characteristics of these two alignments, we present a novel content verification method to improve the system security without much computational overhead. Our experiments on the RSR2015 Part-3 digit-prompted task show that, the DNN based alignment performs on par with the HMM alignment. The results also demonstrate the effectiveness of the proposed Kullback-Leibler (KL) divergence based scoring to reject speech with incorrect pass-phrases.

cs.SD↗

arXiv2018

An Efficient and Accurate Hybrid Method for Simulating Non-Linear Neutrino Structure

We present an efficient and accurate method for simulating massive neutrinos in cosmological structure formation simulations, together with an easy to use public implementation. Our method builds on our earlier implementation of the linear response approximation (LRA) for neutrinos, coupled with an N-body code for cold dark matter particles. The LRA's good behaviour at early times and in the linear regime is preserved, while better following the non-linear clustering of neutrinos on small scales. Massive neutrinos are split into initially "fast" and "slow" components. The fast component is followed analytically with the LRA all the way to redshift zero. The slow component is evolved with the LRA only down to a switch-on redshift $z_ν= 1$, below which it is followed with the particle method, in order to fully account for its non-linear evolution. The slow neutrino particles are initialized at $z = 99$ in order to have accurate positions and velocities at the switch-on time, but are not used to compute the potential until $z \leq 1$, thus avoiding the worst effect of particle shot noise. We show that our hybrid method matches (and for small neutrino masses, exceeds) the accuracy of neutrino particle simulations with substantially lower particle load requirements.

astro-ph.CO↗

arXiv2018

Bayesian model-based spatiotemporal survey design for log-Gaussian Cox process

In geostatistics, the design for data collection is central for accurate prediction and parameter inference. One important class of geostatistical models is log-Gaussian Cox process (LGCP) which is used extensively, for example, in ecology. However, there are no formal analyses on optimal designs for LGCP models. In this work, we develop a novel model-based experimental design for LGCP modeling of spatiotemporal point process data. We propose a new spatially balanced rejection sampling design which directs sampling to spatiotemporal locations that are a priori expected to provide most information. We compare the rejection sampling design to traditional balanced and uniform random designs using the average predictive variance loss function and the Kullback-Leibler divergence between prior and posterior for the LGCP intensity function. Our results show that the rejection sampling method outperforms the corresponding balanced and uniform random sampling designs for LGCP whereas the latter work better for models with Gaussian models. We perform a case study applying our new sampling design to plan a survey for species distribution modeling on larval areas of two commercially important fish stocks on Finnish coastal areas. The case study results show that rejection sampling designs give considerable benefit compared to traditional designs. Results show also that best performing designs may vary considerably between target species.

stat.ME↗