Search arXiv⌕ Search

arXiv subjects

Tao Zhang

Publications and source records attributed to Tao Zhang.

At least 595 records · Page 33Linked to original sources
arXiv2020

On the Turán number of 1-subdivision of $K_{3,t}$

For a graph $H$, the 1-subdivision of $H$, denoted by $H'$, is the graph obtained by replacing the edges of $H$ by internally disjoint paths of length 2. Recently, Conlon, Janzer and Lee (arXiv: 1903.10631) asked the following question: For any integer $s\ge2$, estimate the smallest $t$ such that $\textup{ex}(n,K_{s,t}')=Ω(n^{\frac{3}{2}-\frac{1}{2s}})$. In this paper, we consider the case $s=3$. More precisely, we provide an explicit construction giving \begin{align*} \text{ex}(n,K_{3,30}')=Ω(n^{\frac{4}{3}}), \end{align*} which reduces the estimation for the smallest value of $t$ from a magnitude of $10^{56}$ to the number $30$. The construction is algebraic, which is based on some equations over finite fields.

math.CO↗
arXiv2020

Robust Initial Alignment for SINS/DVL Based on Reconstructed Observation Vectors

Misalignment angle will result in a considerable error for the integration of Doppler Velocity Log (DVL) and of Strapdown Inertial Navigation System (SINS). In this paper, a robust initial alignment method for SINS/DVL is proposed to solve a practical applicable issue, which is that the outputs of DVL are often corrupted by the outliers. Firstly, the alignment principle for SINS/DVL is summarized. Secondly, based on the principle of this alignment method, the apparent velocity model is investigated, and the parameters expression of the apparent velocity model are derived detailed. Using the apparent velocity model, the unknown parameters of the apparent velocity model are estimated by the developed Robust Kalman Filter (RKF), then the reconstructed observation vector, where the outliers are detected and isolated, is reconstructed by the estimated parameters. Based on the reconstructed observation vectors, the initial attitude is determined. Finally, the simulation and field tests are carried out to verify the performance of the proposed method. The test results are shown that the proposed method can detect and isolate the outliers effectively and get better performance than the previous work.

eess.SP↗
arXiv2020

On Hom-Lie antialgebra

In this paper, we introduced the notion of Hom-Lie antialgebras. The representations and cohomology theory of Hom-Lie antialgebras are investigated. We prove that the equivalent classes of abelian extensions of Hom-Lie antialgebras are in one-to-one correspondence to elements of the second cohomology group. We also prove that 1-parameter infinitesimal deformation of a Hom-Lie antialgebra are characterized by 2-cocycles of this Hom-Lie antialgebra with adjoint representation in itself. The notion of Nijenhuis operators of Hom-Lie antialgebra is introduced to describe trivial deformations.

math.RA↗
arXiv2019

Bubble formation due to capillary instability during evaporation of a porous medium

We show that during evaporation of a pore network, liquid can refill the gas occupied pores, snapping off a gas bubble, which then moves to a stable configuration. This phenomenon is induced by the capillary instability due to the wettability heterogeneity of the pore network and has a much smaller time scale as compared to the evaporation process. The capillary instability induced liquid refilling and bubble movement are explained in detail based on the analysis of the images obtained from the visualization experiment. The capillary valve effect, which hinders the movement of the gas-liquid interface and is induced by the sudden geometrical expansion between small and large pores, can be suppressed by the residual liquid in the large pore. For better understanding of the capillary instability induced gas-liquid two-phase transport during evaporation, a novel pore network model is developed, which considers not only the capillary and viscous forces but also the inertial forces that are seldom taken into account in the previous models. The pore network modeling results are in good agreement with the experimental data, demonstrating the effectiveness of the developed pore network model, which opens up a new route for better understanding of the role of inertial forces in two-phase transport in porous media.

cond-mat.soft↗
arXiv2019

Pore network model of evaporation in porous media with continuous and discontinuous corner films

During evaporation in porous media, two types of corner films are distinguished. A continuous corner film is connected to the bulk liquid, while a discontinuous one is not. To disclose their effects on evaporation in porous media, a pore network model with both continuous and discontinuous corner films is developed, which considers the capillary and viscous forces as well as the effects of corner films on the threshold pressures of pores. The capillary valve effect induced by the sudden geometrical expansion between the small and large pores is also taken into account in the model. The developed pore network model agrees well with the evaporation experiment with a quasi 2D micro model porous medium, in terms of not only the variation of the liquid saturation in each pore but also the variation of the total evaporation rate. The pore network models that neglect the corner films or the liquid viscosity are also compared with the experiment so as to shed light on the roles of the corner films. The continuous corner films, which contribute to sustain the high evaporation rate, can be interrupted to be the discontinuous ones not only by the gas invasion into pores but also by the capillary scissors effect due to the local convex topology of the solid matrix.

physics.flu-dyn↗
arXiv2019

Design and commissioning of a 600 L Time Projection Chamber with Microbulk Micromegas

We report the design, construction, and initial commissioning results of a large high pressure gaseous Time Projection Chamber (TPC) with Micromegas modules for charge readout. The detector vessel has an inner volume of about 600 L and an active volume of 270 L. At 10 bar operating pressure, the active volume contains about 20 kg of xenon gas and can image charged particle tracks. Drift electrons are collected by the charge readout plane, which accommodates a tessellation of seven Micromegas modules. Each of the Micromegas covers a square of 20 cm by 20 cm. A new type of Microbulk Micromegas is chosen for this application due to its good gain uniformity and low radioactive contamination. Initial commissioning results with 1 Micromegas module running with 1 bar argon and isobutane gas mixture and 5 bar xenon and trimethylamine (TMA) gas mixture are reported. We also recorded extended background tracks from cosmic ray events and highlighted the unique tracking feature of this gaseous TPC.

physics.ins-det↗
arXiv2019

Visual Tactile Fusion Object Clustering

Object clustering, aiming at grouping similar objects into one cluster with an unsupervised strategy, has been extensivelystudied among various data-driven applications. However, most existing state-of-the-art object clustering methods (e.g., single-view or multi-view clustering methods) only explore visual information, while ignoring one of most important sensing modalities, i.e., tactile information which can help capture different object properties and further boost the performance of object clustering task. To effectively benefit both visual and tactile modalities for object clustering, in this paper, we propose a deep Auto-Encoder-like Non-negative Matrix Factorization framework for visual-tactile fusion clustering. Specifically, deep matrix factorization constrained by an under-complete Auto-Encoder-like architecture is employed to jointly learn hierarchical expression of visual-tactile fusion data, and preserve the local structure of data generating distribution of visual and tactile modalities. Meanwhile, a graph regularizer is introduced to capture the intrinsic relations of data samples within each modality. Furthermore, we propose a modality-level consensus regularizer to effectively align thevisual and tactile data in a common subspace in which the gap between visual and tactile data is mitigated. For the model optimization, we present an efficient alternating minimization strategy to solve our proposed model. Finally, we conduct extensive experiments on public datasets to verify the effectiveness of our framework.

cs.LG↗
arXiv2019

DARB: A Density-Aware Regular-Block Pruning for Deep Neural Networks

The rapidly growing parameter volume of deep neural networks (DNNs) hinders the artificial intelligence applications on resource constrained devices, such as mobile and wearable devices. Neural network pruning, as one of the mainstream model compression techniques, is under extensive study to reduce the number of parameters and computations. In contrast to irregular pruning that incurs high index storage and decoding overhead, structured pruning techniques have been proposed as the promising solutions. However, prior studies on structured pruning tackle the problem mainly from the perspective of facilitating hardware implementation, without analyzing the characteristics of sparse neural networks. The neglect on the study of sparse neural networks causes inefficient trade-off between regularity and pruning ratio. Consequently, the potential of structurally pruning neural networks is not sufficiently mined. In this work, we examine the structural characteristics of the irregularly pruned weight matrices, such as the diverse redundancy of different rows, the sensitivity of different rows to pruning, and the positional characteristics of retained weights. By leveraging the gained insights as a guidance, we first propose the novel block-max weight masking (BMWM) method, which can effectively retain the salient weights while imposing high regularity to the weight matrix. As a further optimization, we propose a density-adaptive regular-block (DARB) pruning that outperforms prior structured pruning work with high pruning ratio and decoding efficiency. Our experimental results show that DARB can achieve 13$\times$ to 25$\times$ pruning ratio, which are 2.8$\times$ to 4.3$\times$ improvements than the state-of-the-art counterparts on multiple neural network models and tasks. Moreover, DARB can achieve 14.3$\times$ decoding efficiency than block pruning with higher pruning ratio.

cs.LG↗
arXiv2019

Some New Results on Splitter Sets

Splitter sets have been widely studied due to their applications in flash memories, and their close relations with lattice tilings and conflict avoiding codes. In this paper, we give necessary and sufficient conditions for the existence of nonsingular perfect splitter sets, $B[-k_1,k_2](p)$ sets, where $0\le k_{1}\leq k_{2}=4$. Meanwhile, constructions of nonsingular perfect splitter sets are given. When perfect splitter sets do not exist, we present four new constructions of quasi-perfect splitter sets. Finally, we give a connection between nonsingular splitter sets and Cayley graphs, and as a byproduct, a general lower bound on the maximum size of nonsingular splitter sets is given.

cs.IT↗
arXiv2019

3-uniform hypergraphs with few Berge paths of length three between any two vertices

Recently, Berge theta hypergraphs have received special attention due to the similarity with Berge even cycles. Let $r$-uniform Berge theta hypergraph $Θ_{\ell,t}^{B}$ be the $r$-uniform hypergraph consisting of $t$ internally disjoint Berge paths of length $\ell$ with the same pair of endpoints. In this work, we determine the Turán number of $3$-uniform Berge theta hypergraph when $\ell=3$ and $t$ is relatively small. More precisely, we provide an explicit construction giving \begin{align*} \textup{ex}_{3}(n,Θ_{3,217}^{B})=Ω(n^{\frac{4}{3}}). \end{align*} This matches an earlier upper bound by He and Tait up to an absolute constant factor. The construction is algebraic, which is based on some equations over finite fields, and the parameter $t$ in our construction is much smaller than that in random algebraic construction. Our main technique is using the resultant of polynomials, which appears to be a powerful technique to eliminate variables.

math.CO↗
arXiv2019

New lower bounds for partial $k$-parallelisms

Due to the applications in network coding, subspace codes and designs have received many attentions. Suppose that $k\mid n$ and $V(n,q)$ is an $n$-dimensional space over the finite field $\mathbb{F}_{q}$. A $k$-spread is a $\frac{q^n-1}{q^k-1}$-set of $k$-dimensional subspaces of $V(n,q)$ such that each nonzero vector is covered exactly once. A partial $k$-parallelism in $V(n,q)$ is a set of pairwise disjoint $k$-spreads. As the number of $k$-dimensional subspaces in $V(n,q)$ is ${n \brack k}_{q}$, there are at most ${n-1 \brack k-1}_{q}$ spreads in a partial $k$-parallelism. By studying the independence numbers of Cayley graphs associated to a special type of partial $k$-parallelisms in $V(n,q)$, we obtain new lower bounds for partial $k$-parallelisms. In particular, we show that there exist at least $\frac{q^{k}-1}{q^{n}-1}{n-1 \brack k-1}_q$ pairwise disjoint $k$-spreads in $V(n,q)$.

math.CO↗
arXiv2019

Higher omni-Lie algebroids

We propose a definition of a "higher" version of the omni-Lie algebroid and study its isotropic and involutive subbundles. Our higher omni-Lie algebroid is to (multi)contact and related geometries what the higher generalized tangent bundle of Zambon and Bi/Sheng is to (multi)symplectic and related geometries.

math.DG↗
arXiv2019

Representation Learning with Ordered Relation Paths for Knowledge Graph Completion

Incompleteness is a common problem for existing knowledge graphs (KGs), and the completion of KG which aims to predict links between entities is challenging. Most existing KG completion methods only consider the direct relation between nodes and ignore the relation paths which contain useful information for link prediction. Recently, a few methods take relation paths into consideration but pay less attention to the order of relations in paths which is important for reasoning. In addition, these path-based models always ignore nonlinear contributions of path features for link prediction. To solve these problems, we propose a novel KG completion method named OPTransE. Instead of embedding both entities of a relation into the same latent space as in previous methods, we project the head entity and the tail entity of each relation into different spaces to guarantee the order of relations in the path. Meanwhile, we adopt a pooling strategy to extract nonlinear and complex features of different paths to further improve the performance of link prediction. Experimental results on two benchmark datasets show that the proposed model OPTransE performs better than state-of-the-art methods.

cs.LG↗
arXiv2019

Channel Splitting Network for Single MR Image Super-Resolution

High resolution magnetic resonance (MR) imaging is desirable in many clinical applications due to its contribution to more accurate subsequent analyses and early clinical diagnoses. Single image super resolution (SISR) is an effective and cost efficient alternative technique to improve the spatial resolution of MR images. In the past few years, SISR methods based on deep learning techniques, especially convolutional neural networks (CNNs), have achieved state-of-the-art performance on natural images. However, the information is gradually weakened and training becomes increasingly difficult as the network deepens. The problem is more serious for medical images because lacking high quality and effective training samples makes deep models prone to underfitting or overfitting. Nevertheless, many current models treat the hierarchical features on different channels equivalently, which is not helpful for the models to deal with the hierarchical features discriminatively and targetedly. To this end, we present a novel channel splitting network (CSN) to ease the representational burden of deep models. The proposed CSN model divides the hierarchical features into two branches, i.e., residual branch and dense branch, with different information transmissions. The residual branch is able to promote feature reuse, while the dense branch is beneficial to the exploration of new features. Besides, we also adopt the merge-and-run mapping to facilitate information integration between different branches. Extensive experiments on various MR images, including proton density (PD), T1 and T2 images, show that the proposed CSN model achieves superior performance over other state-of-the-art SISR methods.

cs.CV↗
arXiv2019

Stabilizing effect of enhanced resistivity on peeling-ballooning instabilities on EAST

Previous stability analysis of NSTX equilibrium with lithium-conditioning demonstrates that the enhanced resistivity due to the increased effective charge number Zeff (i.e. increased impurity level) can provide a stabilizing effect on low-n edge localized modes (Banerjee et al 2017 Nucl. Fusion 24 054501). This paper extends the resistivity stabilizing effect to the intermediate-n peeling-ballooning (PB) instabilities with the linear stability analysis of EAST high-confinement mode equilibria in NIMROD two-fluid calculations. However, the resistivity stabilizing effect on PB instabilities in the EAST tokamak appears weaker than that found in NSTX. This work may give better insight into the physical mechanism behind the beneficial effects of impurity on the pedestal stability.

physics.plasm-ph↗
arXiv2019

A Hybrid Approach with Optimization and Metric-based Meta-Learner for Few-Shot Learning

Few-shot learning aims to learn classifiers for new classes with only a few training examples per class. Most existing few-shot learning approaches belong to either metric-based meta-learning or optimization-based meta-learning category, both of which have achieved successes in the simplified "$k$-shot $N$-way" image classification settings. Specifically, the optimization-based approaches train a meta-learner to predict the parameters of the task-specific classifiers. The task-specific classifiers are required to be homogeneous-structured to ease the parameter prediction, so the meta-learning approaches could only handle few-shot learning problems where the tasks share a uniform number of classes. The metric-based approaches learn one task-invariant metric for all the tasks. Even though the metric-learning approaches allow different numbers of classes, they require the tasks all coming from a similar domain such that there exists a uniform metric that could work across tasks. In this work, we propose a hybrid meta-learning model called Meta-Metric-Learner which combines the merits of both optimization- and metric-based approaches. Our meta-metric-learning approach consists of two components, a task-specific metric-based learner as a base model, and a meta-learner that learns and specifies the base model. Thus our model is able to handle flexible numbers of classes as well as generate more generalized metrics for classification across tasks. We test our approach in the standard "$k$-shot $N$-way" few-shot learning setting following previous works and a new realistic few-shot setting with flexible class numbers in both single-source form and multi-source forms. Experiments show that our approach can obtain superior performance in all settings.

cs.LG↗
arXiv2019

Mixed-Supervised Dual-Network for Medical Image Segmentation

Deep learning based medical image segmentation models usually require large datasets with high-quality dense segmentations to train, which are very time-consuming and expensive to prepare. One way to tackle this challenge is by using the mixed-supervised learning framework, in which only a part of data is densely annotated with segmentation label and the rest is weakly labeled with bounding boxes. The model is trained jointly in a multi-task learning setting. In this paper, we propose Mixed-Supervised Dual-Network (MSDN), a novel architecture which consists of two separate networks for the detection and segmentation tasks respectively, and a series of connection modules between the layers of the two networks. These connection modules are used to transfer useful information from the auxiliary detection task to help the segmentation task. We propose to use a recent technique called "Squeeze and Excitation" in the connection module to boost the transfer. We conduct experiments on two medical image segmentation datasets. The proposed MSDN model outperforms multiple baselines.

cs.CV↗