arXiv subjects

Dong Wang

Publications and source records attributed to Dong Wang.

At least 595 records · Page 33Linked to original sources

arXiv2020

Multi-modal Visual Tracking: Review and Experimental Comparison

Visual object tracking, as a fundamental task in computer vision, has drawn much attention in recent years. To extend trackers to a wider range of applications, researchers have introduced information from multiple modalities to handle specific scenes, which is a promising research prospect with emerging methods and benchmarks. To provide a thorough review of multi-modal track-ing, we summarize the multi-modal tracking algorithms, especially visible-depth (RGB-D) tracking and visible-thermal (RGB-T) tracking in a unified taxonomy from different aspects. Second, we provide a detailed description of the related benchmarks and challenges. Furthermore, we conduct extensive experiments to analyze the effectiveness of trackers on five datasets: PTB, VOT19-RGBD, GTOT, RGBT234, and VOT19-RGBT. Finally, we discuss various future directions from different perspectives, including model design and dataset construction for further research.

cs.CV↗

arXiv2020

Generating Large Convex Polytopes Directly on Point Clouds

In this paper, we present a method to efficiently generate large, free, and guaranteed convex space among arbitrarily cluttered obstacles. Our method operates directly on point clouds, avoids expensive calculations, and processes thousands of points within a few milliseconds, which extremely suits embedded platforms. The base stone of our method is sphere flipping, a one-one invertible nonlinear transformation, which maps a set of unordered points to a nonlinear space. With these wrapped points, we obtain a collision-free star convex polytope. Then, utilizing the star convexity, we efficiently modify the polytope to convex and guarantee its free of obstacles. Extensive quantitative evaluations show that our method significantly outperforms state-of-the-art works in efficiency. We also present practical applications with our method in 3D, including large-scale deformable topological mapping and quadrotor optimal trajectory planning, to validate its capability and efficiency. The source code of our method will be released for the reference of the community.

cs.RO↗

arXiv2020

Can We Trust Deep Speech Prior?

Recently, speech enhancement (SE) based on deep speech prior has attracted much attention, such as the variational auto-encoder with non-negative matrix factorization (VAE-NMF) architecture. Compared to conventional approaches that represent clean speech by shallow models such as Gaussians with a low-rank covariance, the new approach employs deep generative models to represent the clean speech, which often provides a better prior. Despite the clear advantage in theory, we argue that deep priors must be used with much caution, since the likelihood produced by a deep generative model does not always coincide with the speech quality. We designed a comprehensive study on this issue and demonstrated that based on deep speech priors, a reasonable SE performance can be achieved, but the results might be suboptimal. A careful analysis showed that this problem is deeply rooted in the disharmony between the flexibility of deep generative models and the nature of the maximum-likelihood (ML) training.

cs.SD↗

arXiv2020

Deep Normalization for Speaker Vectors

Deep speaker embedding has demonstrated state-of-the-art performance in speaker recognition tasks. However, one potential issue with this approach is that the speaker vectors derived from deep embedding models tend to be non-Gaussian for each individual speaker, and non-homogeneous for distributions of different speakers. These irregular distributions can seriously impact speaker recognition performance, especially with the popular PLDA scoring method, which assumes homogeneous Gaussian distribution. In this paper, we argue that deep speaker vectors require deep normalization, and propose a deep normalization approach based on a novel discriminative normalization flow (DNF) model. We demonstrate the effectiveness of the proposed approach with experiments using the widely used SITW and CNCeleb corpora. In these experiments, the DNF-based normalization delivered substantial performance gains and also showed strong generalization capability in out-of-domain tests.

eess.AS↗

arXiv2020

Remarks on Optimal Scores for Speaker Recognition

In this article, we first establish the theory of optimal scores for speaker recognition. Our analysis shows that the minimum Bayes risk (MBR) decisions for both the speaker identification and speaker verification tasks can be based on a normalized likelihood (NL). When the underlying generative model is a linear Gaussian, the NL score is mathematically equivalent to the PLDA likelihood ratio, and the empirical scores based on cosine distance and Euclidean distance can be seen as approximations of this linear Gaussian NL score under some conditions. We discuss a number of properties of the NL score and perform a simple simulation experiment to demonstrate the properties of the NL score.

cs.LG↗

arXiv2020

Deep generative LDA

Linear discriminant analysis (LDA) is a popular tool for classification and dimension reduction. Limited by its linear form and the underlying Gaussian assumption, however, LDA is not applicable in situations where the data distribution is complex. Recently, we proposed a discriminative normalization flow (DNF) model. In this study, we reinterpret DNF as a deep generative LDA model, and study its properties in representing complex data. We conducted a simulation experiment and a speaker recognition experiment. The results show that DNF and its subspace version are much more powerful than the conventional LDA in modeling complex data and retrieving low-dimensional representations.

cs.LG↗

arXiv2020

Deep Speaker Vector Normalization with Maximum Gaussianality Training

Deep speaker embedding represents the state-of-the-art technique for speaker recognition. A key problem with this approach is that the resulting deep speaker vectors tend to be irregularly distributed. In previous research, we proposed a deep normalization approach based on a new discriminative normalization flow (DNF) model, by which the distributions of individual speakers are arguably transformed to homogeneous Gaussians. This normalization was demonstrated to be effective, but despite this remarkable success, we empirically found that the latent codes produced by the DNF model are generally neither homogeneous nor Gaussian, although the model has assumed so. In this paper, we argue that this problem is largely attributed to the maximum-likelihood (ML) training criterion of the DNF model, which aims to maximize the likelihood of the observations but not necessarily improve the Gaussianality of the latent codes. We therefore propose a new Maximum Gaussianality (MG) training approach that directly maximizes the Gaussianality of the latent codes. Our experiments on two data sets, SITW and CNCeleb, demonstrate that our new MG training approach can deliver much better performance than the previous ML training, and exhibits improved domain generalizability, particularly with regard to cosine scoring.

cs.SD↗

arXiv2020

Deep generative factorization for speech signal

Various information factors are blended in speech signals, which forms the primary difficulty for most speech information processing tasks. An intuitive idea is to factorize speech signal into individual information factors (e.g., phonetic content and speaker trait), though it turns out to be highly challenging. This paper presents a speech factorization approach based on a novel factorial discriminative normalization flow model (factorial DNF). Experiments conducted on a two-factor case that involves phonetic content and speaker trait demonstrates that the proposed factorial DNF has powerful capability to factorize speech signals and outperforms several comparative models in terms of information representation and manipulation.

cs.SD↗

arXiv2020

Squeezing value of cross-domain labels: a decoupled scoring approach for speaker verification

Domain mismatch often occurs in real applications and causes serious performance reduction on speaker verification systems. The common wisdom is to collect cross-domain data and train a multi-domain PLDA model, with the hope to learn a domain-independent speaker subspace. In this paper, we firstly present an empirical study to show that simply adding cross-domain data does not help performance in conditions with enrollment-test mismatch. Careful analysis shows that this striking result is caused by the incoherent statistics between the enrollment and test conditions. Based on this analysis, we present a decoupled scoring approach that can maximally squeeze the value of cross-domain labels and obtain optimal verification scores when the enrollment and test are mismatched. When the statistics are coherent, the new formulation falls back to the conventional PLDA. Experimental results on cross-channel test show that the proposed approach is highly effective and is a principle solution to domain mismatch.

cs.SD↗

arXiv2020

An analytic study on the excited states of holographic superconductors

Based on the Sturm-Liouville eigenvalue problem, we develop a general analytic technique to investigate the excited states of the holographic superconductors. By including more higher order terms in the expansion of the trial function, we observe that the analytic results agree well with the numeric data, which indicates that the Sturm-Liouville method is very powerful to study the holographic superconductors even if we consider the excited states. For both the holographic s-wave and p-wave models, we find that the excited state has a lower critical temperature than the corresponding ground state and the difference of the dimensionless critical chemical potential between the consecutive states is around 5. Moreover, we analytically confirm that the holographic superconductor phase transition with the excited states belongs to the second order, which can be used to back up the numerical findings for both s-wave and p-wave superconductors.

hep-th↗

arXiv2020

Consistency of archetypal analysis

Archetypal analysis is an unsupervised learning method that uses a convex polytope to summarize multivariate data. For fixed $k$, the method finds a convex polytope with $k$ vertices, called archetype points, such that the polytope is contained in the convex hull of the data and the mean squared distance between the data and the polytope is minimal. In this paper, we prove a consistency result that shows if the data is independently sampled from a probability measure with bounded support, then the archetype points converge to a solution of the continuum version of the problem, of which we identify and establish several properties. We also obtain the convergence rate of the optimal objective values under appropriate assumptions on the distribution. If the data is independently sampled from a distribution with unbounded support, we also prove a consistency result for a modified method that penalizes the dispersion of the archetype points. Our analysis is supported by detailed computational experiments of the archetype points for data sampled from the uniform distribution in a disk, the normal distribution, an annular distribution, and a Gaussian mixture model.

math.ST↗

arXiv2020

AP20-OLR Challenge: Three Tasks and Their Baselines

This paper introduces the fifth oriental language recognition (OLR) challenge AP20-OLR, which intends to improve the performance of language recognition systems, along with APSIPA Annual Summit and Conference (APSIPA ASC). The data profile, three tasks, the corresponding baselines, and the evaluation principles are introduced in this paper. The AP20-OLR challenge includes more languages, dialects and real-life data provided by Speechocean and the NSFC M2ASR project, and all the data is free for participants. The challenge this year still focuses on practical and challenging problems, with three tasks: (1) cross-channel LID, (2) dialect identification and (3) noisy LID. Based on Kaldi and Pytorch, recipes for i-vector and x-vector systems are also conducted as baselines for the three tasks. These recipes will be online-published, and available for participants to configure LID systems. The baseline results on the three tasks demonstrate that those tasks in this challenge are worth paying more efforts to achieve better performance.

eess.AS↗

arXiv2020

A deep learning approach to multi-track location and orientation in gaseous drift chambers

Accurate measuring the location and orientation of individual particles in a beam monitoring system is of particular interest to researchers in multiple disciplines. Among feasible methods, gaseous drift chambers with hybrid pixel sensors have the great potential to realize long-term stable measurement with considerable precision. In this paper, we introduce deep learning to analyze patterns in the beam projection image to facilitate three-dimensional reconstruction of particle tracks. We propose an end-to-end neural network based on segmentation and fitting for feature extraction and regression. Two segmentation branches, named binary segmentation and semantic segmentation, perform initial track determination and pixel-track association. Then pixels are assigned to multiple tracks, and a weighted least squares fitting is implemented with full back-propagation. Besides, we introduce a center-angle measure to judge the precision of location and orientation by combining two separate factors. The initial position resolution achieves 8.8 $μm$ for the single track and 11.4 $μm$ (15.2 $μm$) for the 1-3 tracks (1-5 tracks), and the angle resolution achieves 0.15$^{\circ}$ and 0.21$^{\circ}$ (0.29$^{\circ}$) respectively. These results show a significant improvement in accuracy and multi-track compatibility compared to traditional methods.

physics.data-an↗

arXiv2020

Newcomer Candidate: Characterizing Contributions of a Novice Developer to GitHub

Context: To attract, onboard, and retain any new-comer in Open Source Software (OSS) projects is vital to their livelihood. Recent studies conclude that OSS projects risk failure due to abandonment and poor participation of newcomers. Evidence suggests more new users are joining GitHub, however, the extent to which they contribute to OSS projects is unknown. Objective: In this study, we coin the term 'newcomer candidate' to describe new users to the GitHub platform. Our objective is to track and characterize their initial contributions. As a preliminary survey, we collected 208 newcomer candidate contributions in GitHub. Using this dataset, we then plan to track their contributions to reveal insights. Method: We will use a mixed-methods approach, i.e., quantitative and qualitative, to identify whether or not newcomer candidates practice social coding, the kinds of their contributions, projects they target, and the proportion that they eventually onboard to an OSS project. Limitation: The key limitation is that our newcomer candidates are restricted to those that were collected from our preliminary survey.

cs.SE↗

arXiv2020

WANA: Symbolic Execution of Wasm Bytecode for Cross-Platform Smart Contract Vulnerability Detection

Many popular blockchain platforms are supporting smart contracts for building decentralized applications. However, the vulnerabilities within smart contracts have led to serious financial loss to their end users. For the EOSIO blockchain platform, effective vulnerability detectors are still limited. Furthermore, existing vulnerability detection tools can only support one blockchain platform. In this work, we present WANA, a cross-platform smart contract vulnerability detection tool based on the symbolic execution of WebAssembly bytecode. Furthermore, WANA proposes a set of test oracles to detect the vulnerabilities in EOSIO and Ethereum smart contracts based on WebAssembly bytecode analysis. Our experimental analysis shows that WANA can effectively detect vulnerabilities in both EOSIO and Ethereum smart contracts with high efficiency.

cs.SE↗

arXiv2020

Energy recovery operation for continuous-wave X-ray free-electron lasers

A superconducting linear accelerator operating in continuous-wave mode could produce X-ray free electron lasers (XFEL) at megahertz repetition rate, with the capability that delivering wide spectral range coherent radiation to multi end stations. In this Letter, the energy recovery Linac (ERL) mode is proposed to flexibly control the electron beam energy for a continuous-wave superconducting Linac. Theoretical investigations and multi-dimensional numerical simulations are applied to the Linac case of Shanghai high-repetition-rate XFEL and extreme light facility. The results show that, with ERL operation in the last 25 cryo-modules, the strict requirements on RF power system could be significantly relaxed. And if one exhaust the RF power, the maximum electron beam energy can be enhanced from 8.74 GeV to 11.41GeV in ERL mode. The optimization of the ERL operation, the multi-energy electron beam transport and the XFEL performance improvements are presented.

physics.acc-ph↗

arXiv2020

Experimental demonstration of complementarity relations between quantum steering criteria

The ability that one system immediately affects another one by using local measurements is regarded as quantum steering, which can be detected by various steering criteria. Recently, Mondal et al. [Phys. Rev. A 98, 052330 (2018)] derived the complementarity relations of coherence steering criteria, and revealed that the quantum steering of system can be observed through the average coherence of subsystem. Here, we experimentally verify the complementarity relations between quantum steering criteria by employing two-photon Bell-like states and three Pauli operators. The results demonstrate that if prepared quantum states can violate two setting coherence steering criteria and turn out to be steerable states, then it cannot violate the complementary settings criteria. Three measurement settings inequality, which establish a complementarity relation between these two coherence steering criteria, always holds in experiment. Besides, we experimentally certify that the strengths of coherence steering criteria dependent on the choice of coherence measure. In comparison with two setting coherence steering criteria based on l1 norm of coherence and relative entropy of coherence, our experimental results show that the steering criterion based on skew information of coherence is more stronger in detecting the steerability of quantum states. Thus, our experimental demonstrations can deepen the understanding of the relation between the quantum steering and quantum coherence.

quant-ph↗

arXiv2020

Multispectral Pan-sharpening via Dual-Channel Convolutional Network with Convolutional LSTM Based Hierarchical Spatial-Spectral Feature Fusion

Multispectral pan-sharpening aims at producing a high resolution (HR) multispectral (MS) image in both spatial and spectral domains by fusing a panchromatic (PAN) image and a corresponding MS image. In this paper, we propose a novel dual-channel network (DCNet) framework for MS pan-sharpening. In our DCNet, the dual-channel backbone involves a spatial channel to capture spatial information with a 2D CNN, and a spectral channel to extract spectral information with a 3D CNN. This heterogeneous 2D/3D CNN architecture can minimize causing spectral information distortion, which typically happens in conventional 2D CNN models. In order to fully integrate the spatial and spectral features captured from different levels, we introduce a multi-level fusion strategy. Specifically, a spatial-spectral CLSTM (S$^2$-CLSTM) module is proposed for fusing the hierarchical spatial and spectral features, which can effectively capture correlations among multi-level features. The S$^2$-CLSTM module attaches two fusion ways: the intra-level fusion via bi-directional lateral connections and inter-level fusion via the cell state in the S$^2$-CLSTM. Finally, the ideal HR-MS image is recovered by a reconstruction module. Extensive experiments have been conducted at both simulated lower scale and the original scale of real-world datasets. Compared with the state-of-the-art methods, the proposed DCNet achieves superior or competitive performance.

eess.IV↗