Searcharxiv⌕ Search

arXiv subjects

Tao Yu

Publications and source records attributed to Tao Yu.

375 records · Page 21Linked to original sources
arXiv2018

A Guide of Fingerprint Based Radio Emitter Localization using Multiple Sensors

Location information is essential to varieties of applications. It is one of the most important context to be detected by wireless distributed sensors, which is a key technology in Internet-of-Things. Fingerprint-based methods, which compare location unique fingerprints collected beforehand with the fingerprint measured from the target, have attracted much attention recently in both of academia and industry. They have been successfully used for many location-based applications.From the viewpoint of practical applications, in this paper, four different typical approaches of fingerprint-based radio emitter localization system are introduced with four different representative applications: localization of LTE smart phone used for anti-cheating in exams, indoor localization of Wi-Fi terminals, localized light control in BEMS using location information of occupants, and illegal radio localization in outdoor environments. Based on the different practical application scenarios, different solutions, which are designed to enhance the localization performance, are discussed in detail. To the best of the authors' knowledge, this is the first paper to give a guideline for readers about fingerprint-based localization system in terms of fingerprint selection, hardware architecture design and algorithm enhancement

eess.SP↗
arXiv2018

Curvature-based Comparison of Two Neural Networks

In this paper we show the similarities and differences of two deep neural networks by comparing the manifolds composed of activation vectors in each fully connected layer of them. The main contribution of this paper includes 1) a new data generating algorithm which is crucial for determining the dimension of manifolds; 2) a systematic strategy to compare manifolds. Especially, we take Riemann curvature and sectional curvature as part of criterion, which can reflect the intrinsic geometric properties of manifolds. Some interesting results and phenomenon are given, which help in specifying the similarities and differences between the features extracted by two networks and demystifying the intrinsic mechanism of deep neural networks.

cs.LG↗
arXiv2017

The Local Dimension of Deep Manifold

Based on our observation that there exists a dramatic drop for the singular values of the fully connected layers or a single feature map of the convolutional layer, and that the dimension of the concatenated feature vector almost equals the summation of the dimension on each feature map, we propose a singular value decomposition (SVD) based approach to estimate the dimension of the deep manifolds for a typical convolutional neural network VGG19. We choose three categories from the ImageNet, namely Persian Cat, Container Ship and Volcano, and determine the local dimension of the deep manifolds of the deep layers through the tangent space of a target image. Through several augmentation methods, we found that the Gaussian noise method is closer to the intrinsic dimension, as by adding random noise to an image we are moving in an arbitrary dimension, and when the rank of the feature matrix of the augmented images does not increase we are very close to the local dimension of the manifold. We also estimate the dimension of the deep manifold based on the tangent space for each of the maxpooling layers. Our results show that the dimensions of different categories are close to each other and decline quickly along the convolutional layers and fully connected layers. Furthermore, we show that the dimensions decline quickly inside the Conv5 layer. Our work provides new insights for the intrinsic structure of deep neural networks and helps unveiling the inner organization of the black box of deep neural networks.

cs.CV↗
arXiv2017

Leveraging Sparse and Dense Feature Combinations for Sentiment Classification

Neural networks are one of the most popular approaches for many natural language processing tasks such as sentiment analysis. They often outperform traditional machine learning models and achieve the state-of-art results on most tasks. However, many existing deep learning models are complex, difficult to train and provide a limited improvement over simpler methods. We propose a simple, robust and powerful model for sentiment classification. This model outperforms many deep learning models and achieves comparable results to other deep learning models with complex architectures on sentiment analysis datasets. We publish the code online.

cs.CL↗
arXiv2017

Design and implementation of lighting control system using battery-less wireless human detection sensor networks

Artificial lighting is responsible for a large portion of total energy consumption and has great potential for energy saving. This paper designs an LED light control algorithm based on users' localization using multiple battery-less binary human detection sensors. The proposed lighting control system focuses on reducing office lighting energy consumption and satisfying users' illumination requirement. Most current lighting control systems use infrared human detection sensors, but the poor detection probability, especially for a static user, makes it difficult to realize comfortable and effective lighting control. To improve the detection probability of each sensor, we proposed to locate sensors as close to each user as possible by using a battery-less wireless sensor network, in which all sensors can be placed freely in the space with high energy stability. We also proposed to use a multi-sensor-based user localization algorithm to capture user's position more accurately and realize fine lighting control which works even with static users. The system is actually implemented in an indoor office environment in a pilot project. A verification experiment is conducted by measuring the practical illumination and power consumption. The performance agrees with design expectations. It shows that the proposed LED lighting control system reduces the energy consumption significantly, 57% compared to the batch control scheme, and satisfies user's illumination requirement with 100% probability.

eess.SY↗
arXiv2016

Sensitivity, proximal extension and higher order almost automorphy

Let $(X,T)$ be a topological dynamical system, and $\mathcal{F}$ be a family of subsets of $\mathbb{Z}_+$. $(X,T)$ is strongly $\mathcal{F}$-sensitive, if there is $δ>0$ such that for each non-empty open subset $U$, there are $x,y\in U$ with $\{n\in\mathbb{Z}_+: d(T^nx,T^ny)>δ\}\in\mathcal{F}$. Let $\mathcal{F}_t$ (resp. $\mathcal{F}_{ip}$, $\mathcal{F}_{fip}$) be consisting of thick sets (resp. IP-sets, subsets containing arbitrarily long finite IP-sets). The following Auslander-Yorke's type dichotomy theorems are obtained: (1) a minimal system is either strongly $\mathcal{F}_{fip}$-sensitive or an almost one-to-one extension of its $\infty$-step nilfactor. (2) a minimal system is either strongly $\mathcal{F}_{ip}$-sensitive or an almost one-to-one extension of its maximal distal factor. (3) a minimal system is either strongly $\mathcal{F}_{t}$-sensitive or a proximal extension of its maximal distal factor.

math.DS↗
arXiv2016

Proposal of Readout Electronics for CSNS-WNS BaF2 Detector

BaF2 (Barium fluoride) detector is one of the experiment facilities at the under construction CSNS-WNS (White Neutron Source at China Spallation Neutron Source). It is designed for precisely measuring (n,gamma) cross section with total 92 crystal elements and completely 4 pi steradian coverage. In this proposal for readout electronics, waveform digitizing technique with 1GSps sampling rate and 12-bit resolution is adopted to precisely capture the detector signal. To solve the problem of massive data readout and processing, the readout electronics system is designed into a distributed architecture with 4 PXIe crates. The digitized detector's signal is concentrated to PXIe crate controller through PCIe bus on backplane and transmitted to data acquisition system over Gigabit Ethernet in parallel. Besides, clock and trigger can be fanned out synchronously to each electronic channel over a high-precision distributing network. Test results showed that the prototype of the readout electronics system achieved good performance and cooperated well.

physics.ins-det↗
arXiv2016

Square-free Discriminants of Matrices and the Generalized Spectral Characterizations of Graphs

Let $S_n(\mathbb{Z})$ and $O_n(\mathbb{Q})$ denote the set of all $n\times n$ symmetric matrices over the ring of integers $\mathbb{Z}$ and the set of all $n\times n$ orthogonal matrices over the field of rational numbers $\mathbb{Q}$, respectively. The paper is mainly concerned with the following problem: Given a matrix $A\in {S_n(\mathbb{Z})}$. How can one find all rational orthogonal matrices $Q\in{O_n(\mathbb{Q})}$ such that $Q^TAQ\in {S_n(\mathbb{Z})}$, and in particular, when does $Q^TAQ\in {S_n(\mathbb{Z})}$ with $Q\in{O_n(\mathbb{Q})}$ imply that $Q$ is \emph{a signed permutation matrix} (i.e., the matrix obtained from a permutation matrix $P$ by replacing each 1 in $P$ with 1 or $-1$)? A surprisingly simple answer was given in terms of whether the discriminant of the characteristic polynomial of $A$ is odd and square-free, which partially answers the above questions. More precisely, let $Δ_A=\pm \res(ϕ,ϕ')$ be \emph{the discriminant of matrix $A$}, where $\res(ϕ,ϕ')$ is \emph{the resultant} of the characteristic polynomial $ϕ$ of $A$ and its derivative $ϕ'$. We show that if $Δ_A$ is odd and square-free, then $Q^TAQ\in {S_n(\mathbb{Z})}$ with $Q\in{O_n(\mathbb{Q})}$ implies that $Q$ is a signed permutation matrix. As an application, we present a simple and efficient method for testing whether a graph is determined by the generalized spectrum, which significantly extends our previous work.

math.CO↗
arXiv2016

Statistical Theory of Initiation of Explosives by Impact

When a given weight dropped onto an explosive charge, explosion or not is probabilistic for certain impact energy and the frequency of explosion is always increase with increasing impact energy. Based on experimental results and recently theoretical work, we propose that the hot spot formation is attributed to the activated molecules decomposition and the number of molecules initiation is proportional to the impact energy but not the dropped weight heating as the previous hot spot theory. A theoretical model based on two states model has been put forward for this phenomena. It is shown that the activated molecules to form a hot spot determine the probabilistic nature of initiation by impact. It is shown a good agreement tested with Hexogen (RDX) experimental impact data.

cond-mat.soft↗
arXiv2014

Maximum Smoothed Likelihood Component Density Estimation in Mixture Models with Known Mixing Proportions

In this paper, we propose a maximum smoothed likelihood method to estimate the component density functions of mixture models, in which the mixing proportions are known and may differ among observations. The proposed estimates maximize a smoothed log likelihood function and inherit all the important properties of probability density functions. A majorization-minimization algorithm is suggested to compute the proposed estimates numerically. In theory, we show that starting from any initial value, this algorithm increases the smoothed likelihood function and further leads to estimates that maximize the smoothed likelihood function. This indicates the convergence of the algorithm. Furthermore, we theoretically establish the asymptotic convergence rate of our proposed estimators. An adaptive procedure is suggested to choose the bandwidths in our estimation procedure. Simulation studies show that the proposed method is more efficient than the existing method in terms of integrated squared errors. A real data example is further analyzed.

stat.ME↗
arXiv2013

Local tests for identifying anisotropic diffusion areas in human brain with DTI

Diffusion tensor imaging (DTI) plays a key role in analyzing the physical structures of biological tissues, particularly in reconstructing fiber tracts of the human brain in vivo. On the one hand, eigenvalues of diffusion tensors (DTs) estimated from diffusion weighted imaging (DWI) data usually contain systematic bias, which subsequently biases the diffusivity measurements popularly adopted in fiber tracking algorithms. On the other hand, correctly accounting for the spatial information is important in the construction of these diffusivity measurements since the fiber tracts are typically spatially structured. This paper aims to establish test-based approaches to identify anisotropic water diffusion areas in the human brain. These areas in turn indicate the areas passed by fiber tracts. Our proposed test statistic not only takes into account the bias components in eigenvalue estimates, but also incorporates the spatial information of neighboring voxels. Under mild regularity conditions, we demonstrate that the proposed test statistic asymptotically follows a $χ^2$ distribution under the null hypothesis. Simulation and real DTI data examples are provided to illustrate the efficacy of our proposed methods.

stat.AP↗
arXiv2011

Multiple testing via $FDR_L$ for large-scale imaging data

The multiple testing procedure plays an important role in detecting the presence of spatial signals for large-scale imaging data. Typically, the spatial signals are sparse but clustered. This paper provides empirical evidence that for a range of commonly used control levels, the conventional $\operatorname {FDR}$ procedure can lack the ability to detect statistical significance, even if the $p$-values under the true null hypotheses are independent and uniformly distributed; more generally, ignoring the neighboring information of spatially structured data will tend to diminish the detection effectiveness of the $\operatorname {FDR}$ procedure. This paper first introduces a scalar quantity to characterize the extent to which the "lack of identification phenomenon" ($\operatorname {LIP}$) of the $\operatorname {FDR}$ procedure occurs. Second, we propose a new multiple comparison procedure, called $\operatorname {FDR}_L$, to accommodate the spatial information of neighboring $p$-values, via a local aggregation of $p$-values. Theoretical properties of the $\operatorname {FDR}_L$ procedure are investigated under weak dependence of $p$-values. It is shown that the $\operatorname {FDR}_L$ procedure alleviates the $\operatorname {LIP}$ of the $\operatorname {FDR}$ procedure, thus substantially facilitating the selection of more stringent control levels. Simulation evaluations indicate that the $\operatorname {FDR}_L$ procedure improves the detection sensitivity of the $\operatorname {FDR}$ procedure with little loss in detection specificity. The computational simplicity and detection effectiveness of the $\operatorname {FDR}_L$ procedure are illustrated through a real brain fMRI dataset.

math.ST↗
arXiv2008

Semiparametric detection of significant activation for brain fMRI

Functional magnetic resonance imaging (fMRI) aims to locate activated regions in human brains when specific tasks are performed. The conventional tool for analyzing fMRI data applies some variant of the linear model, which is restrictive in modeling assumptions. To yield more accurate prediction of the time-course behavior of neuronal responses, the semiparametric inference for the underlying hemodynamic response function is developed to identify significantly activated voxels. Under mild regularity conditions, we demonstrate that a class of the proposed semiparametric test statistics, based on the local linear estimation technique, follow $χ^2$ distributions under null hypotheses for a number of useful hypotheses. Furthermore, the asymptotic power functions of the constructed tests are derived under the fixed and contiguous alternatives. Simulation evaluations and real fMRI data application suggest that the semiparametric inference procedure provides more efficient detection of activated brain areas than the popular imaging analysis tools AFNI and FSL.

math.ST↗
arXiv2007

Atomic decoration for improving the efficiency of field electron emission of carbon nanotubes

The field electron emission from the single-walled carbon nanotubes with their open ends terminated by -BH, -NH, and -O has been simulated. The apex-vacuum barrier and the emission current have been calculated. It has been found that -BH and -NH suppress the apex-vacuum barrier significantly and lead to higher emission current in contrast to the -O terminated structure in the same applied field. The calculated binding energy implies that the carbon nanotubes terminated with -BH and -NH are more stable than those saturated by oxygen atoms or by hydrogen atoms.

cond-mat.mtrl-sci↗