arXiv subjects

Qiang Wu

Publications and source records attributed to Qiang Wu.

At least 145 records · Page 8Linked to original sources

arXiv2020

Optimal Rates of Distributed Regression with Imperfect Kernels

Distributed machine learning systems have been receiving increasing attentions for their efficiency to process large scale data. Many distributed frameworks have been proposed for different machine learning tasks. In this paper, we study the distributed kernel regression via the divide and conquer approach. This approach has been proved asymptotically minimax optimal if the kernel is perfectly selected so that the true regression function lies in the associated reproducing kernel Hilbert space. However, this is usually, if not always, impractical because kernels that can only be selected via prior knowledge or a tuning process are hardly perfect. Instead it is more common that the kernel is good enough but imperfect in the sense that the true regression can be well approximated by but does not lie exactly in the kernel space. We show distributed kernel regression can still achieves capacity independent optimal rate in this case. To this end, we first establish a general framework that allows to analyze distributed regression with response weighted base algorithms by bounding the error of such algorithms on a single data set, provided that the error bounds has factored the impact of the unexplained variance of the response variable. Then we perform a leave one out analysis of the kernel ridge regression and bias corrected kernel ridge regression, which in combination with the aforementioned framework allows us to derive sharp error bounds and capacity independent optimal rates for the associated distributed kernel regression algorithms. As a byproduct of the thorough analysis, we also prove the kernel ridge regression can achieve rates faster than $N^{-1}$ (where $N$ is the sample size) in the noise free setting which, to our best knowledge, are first observed and novel in regression learning.

cs.LG↗

arXiv2020

Low-Rank Pairwise Alignment Bilinear Network For Few-Shot Fine-Grained Image Classification

Deep neural networks have demonstrated advanced abilities on various visual classification tasks, which heavily rely on the large-scale training samples with annotated ground-truth. However, it is unrealistic always to require such annotation in real-world applications. Recently, Few-Shot learning (FS), as an attempt to address the shortage of training samples, has made significant progress in generic classification tasks. Nonetheless, it is still challenging for current FS models to distinguish the subtle differences between fine-grained categories given limited training data. To filling the classification gap, in this paper, we address the Few-Shot Fine-Grained (FSFG) classification problem, which focuses on tackling the fine-grained classification under the challenging few-shot learning setting. A novel low-rank pairwise bilinear pooling operation is proposed to capture the nuanced differences between the support and query images for learning an effective distance metric. Moreover, a feature alignment layer is designed to match the support image features with query ones before the comparison. We name the proposed model Low-Rank Pairwise Alignment Bilinear Network (LRPABN), which is trained in an end-to-end fashion. Comprehensive experimental results on four widely used fine-grained classification datasets demonstrate that our LRPABN model achieves the superior performances compared to state-of-the-art methods.

cs.CV↗

arXiv2020

Rotating regular black holes in conformal massive gravity

In this paper, we use a suitable conformal rescaling to construct static and rotating regular black holes in conformal massive gravity. The new metric is characterized by the mass $M$, the "scalar charge" $Q$, the angular momentum parameter $a$, the "hair parameter" $λ$, and the conformal scale factor encoded in the parameter $L$. We explore the shadow images and the deflection angles of relativistic massive particles in the spacetime geometry of a rotating regular black hole. For $λ\neq 0$ and $Q > 0$, the shadow is larger than the shadow of a Kerr black hole. In particular, if $λ< 0$, the shadow radius increases considerably. For $λ\neq 0$ and $Q < 0$, the shadow is smaller than the shadow of a Kerr black hole. Additionally we put observational constraints on the parameter $ Q $ using the latest Event Horizon Telescope (EHT) observation of the supermassive black hole M87*. Lastly, using the Gauss-Bonnet theorem, we show that the deflection angle of massive particles is strongly affected by the parameter $L$. The deflection angle might be used to distinguish rotating regular black holes from rotating singular black holes.

gr-qc↗

arXiv2019

Inflationary Cosmology with Quantum Gravitational Effects and the Swampland Conjectures

Recently proposed two swampland criteria that arising from string theory landscape lead to the important challenge of the realization of single-field inflationary models. Especially one of swampland criteria which implies a large tensor-to-scalar ratio is strongly in tension with recent observational results. In this paper, we explore the possibility if the swampland conjectures could be compatible with single-field inflationary scenarios if the effects due to the quantum theory of gravity are considered. We show that the quantum gravitational effects due to the nonlinear dispersion relation provides significant modifications on the amplitude of both the scalar and tensor perturbation spectra. Such modifications could be either raise or reduce the perturbation spectra depending on the values of the parameters in the nonlinear terms of the dispersion relations. Therefore, these effects can reduce the tensor-to-scalar ratio to a smaller value which helps to relax the tension between the swampland conjecture and observational data.

gr-qc↗

arXiv2019

Inflationary perturbation spectra at next-to-leading slow-roll order in effective field theory of inflation

The effective field theory (EFT) of inflation provides an essential picture to explore the effects of the unknown high energy physics in the single scalar field inflation models. For a generic EFT of inflation, possible high energy corrections to simple slow-roll inflation can modify both the propagating speed and dispersion relations of the cosmological scalar and tensor perturbations. With the arrival of the era of precision cosmology, it is expected that these high energy corrections become more important and have to be taken into account in the analysis with future precise observational data. In this paper we study the observational predictions of the EFT of inflation by using the third-order uniform asymptotic approximation method. We calculate explicitly the primordial power spectra, spectral indices, running of the spectral indices for both scalar and tensor perturbations, and the ratio between tensor and scalar spectra. These expressions are all written in terms of the Hubble flow parameters and the flow of four new slow-roll parameters and expanded up to the next-to-leading order in the slow-roll expansions so they represent the most accurate results obtained so far in the literature. The flow of the four new slow-roll parameters, which arise from the four new operators introduced in the action of the EFT of inflation, can affect the primordial perturbation spectra at the leading-order and the corresponding spectral indices at the next-to-leading order.

gr-qc↗

arXiv2019

Photonic hooks from Janus microcylinders

Recently, a type of curved light beams, photonic hooks (PHs), was theoretically predicted and experimentally observed. The production of photonic hook (PH) is due to the breaking of structural symmetry of a plane-wave illuminated microparticle. Herein, we presented and implemented a new approach, of utilizing the symmetry-broken of the microparticles in material composition, for the generation of PHs from Janus microcylinders. Finite element method based numerical simulation and energy flow diagram represented theoretical analysis were used to investigate the field distribution characteristics and formation mechanism of the PHs. The full width at half-maximum (FWHM) of the PH (~0.29$λ$) is smaller than the FWHM of the photonic nanojet (~0.35$λ$) formed from a circular microcylinder with the same geometric radius. By changing the refractive index contrasts between upper and lower half-cylinders, or rotating the Janus microcylinder relative to the central axis, the shape profiles of the PHs can be efficiently modulated. The tunability of the PHs through simple stretching or compression operations, for the Janus microcylinder constituted by one solid inorganic half-cylinder and the other flexible polymer half-cylinder, was studied and discussed as well.

physics.optics↗

arXiv2019

High coherent frequency-entangled photons generated by parametric instability in active fiber ring cavity

High coherent frequency-entangled photons at telecom band are critical in quantum information protocols and quantum tele-communication. While photon pairs generated by spontaneous parametric down-conversion in nonlinear crystal or modulation instability in optical fiber exhibit random fluctuations, making the photons distinguishable among consecutive roundtrips. Here, we demonstrate a frequency-entangled photons based on parametric instability in an active fiber ring cavity, where periodic modulation of dispersion excites parametric resonance. The characteristic wave number in parametric instability is selected by the periodic modulation of resonator, and stable patterns with symmetric gains are formed. We find that the spectra of parametric instability sidebands possess a high degree of coherence, which is verified by the background-free autocorrelation of single-shot spectra. Two photon interference is performed by a fiber-based Mach-Zehnder interferometer without any stabilization. We obtain a Hong-Ou-Mandel interference visibility of 86.3% with a dip width of 4.3 mm. The correlation time measurement exhibits a linewidth of 68.36 MHz, indicating high coherence and indistinguishability among the photon pairs. Our results proves that the parametric instability in active fiber cavity is effective to generate high coherent frequency-entangled photon pairs, which would facilitate subsequent quantum applications.

physics.optics↗

arXiv2019

Inflationary perturbation spectrum in extended effective field theory of inflation

The effective field theory (EFT) of inflation provides a natural framework to study the new physical effects on primordial perturbations. Recently a healthy extension of the EFT of inflation with high-order operators has been proposed, which avoids ghosts and meanwhile leads to a nonlinear dispersion relation of the scalar perturbations. This paper is devoted to studying the effects of these high-order operators by using the uniform asymptotic approximation method. In particular, we first construct the approximate analytical solution to the mode function of the scalar perturbations. Because of the presence of the high-order operators, the perturbation modes usually experience a period of non-adiabatic evolution before they cross the Hubble radius, which could lead to the production of excited states and modifications of the primordial perturbation spectrum. However, we show that the modified power spectrum is still nearly scale-invariant and the presence of the high-order operators can only affect the overall amplitude of the spectrum. In particular, after showing explicitly the impact of these new effects on particle production rate and perturbation spectrum, we explore their origin in detail.

hep-th↗

arXiv2019

SBSGAN: Suppression of Inter-Domain Background Shift for Person Re-Identification

Cross-domain person re-identification (re-ID) is challenging due to the bias between training and testing domains. We observe that if backgrounds in the training and testing datasets are very different, it dramatically introduces difficulties to extract robust pedestrian features, and thus compromises the cross-domain person re-ID performance. In this paper, we formulate such problems as a background shift problem. A Suppression of Background Shift Generative Adversarial Network (SBSGAN) is proposed to generate images with suppressed backgrounds. Unlike simply removing backgrounds using binary masks, SBSGAN allows the generator to decide whether pixels should be preserved or suppressed to reduce segmentation errors caused by noisy foreground masks. Additionally, we take ID-related cues, such as vehicles and companions into consideration. With high-quality generated images, a Densely Associated 2-Stream (DA-2S) network is introduced with Inter Stream Densely Connection (ISDC) modules to strengthen the complementarity of the generated data and ID-related cues. The experiments show that the proposed method achieves competitive performance on three re-ID datasets, ie., Market-1501, DukeMTMC-reID, and CUHK03, under the cross-domain person re-ID scenario.

cs.CV↗

arXiv2019

An analytical approach to the field amplification and particle production by parametric resonance during inflation and reheating

Field amplification and particle production due to parametric resonance are highly nontrivial predictions of quantum fields that couple to an oscillating source during inflation and reheating. Understanding this two effects is crucial for the connection between the resonance phenomenon and precise observational data. In this paper, we give a general and analytic analysis of parametric resonance of relevant field modes evolving during inflation and reheating by using the uniform asymptotic approximation. This analysis can provide a clear and quantitative explanation for the field amplification and particle production during the resonance. The potential applications of our results to several examples, including sound resonance during inflation, particle productions during reheating, and parametric resonance due to self-resonance potentials, have also been explored. The formalism developed in this paper is also applicable to parametric resonance in a broad areas of modern science.

hep-ph↗

arXiv2019

Shadows and deflection angle of charged and slowly rotating black holes in Einstein-Æther theory

In this paper, we study the shadow cast by two types of charged and slowly rotating black holes in Eisntein-aether theory of gravity. This two types of black holes are corresponding to two specific combinations of the coupling constants of the aether field, i.e., $c_{14} = 0$ but $c_{123}\neq 0$ for the first type and $c_{123} = 0$ for the second type, respectively. For both types of black holes, in addition to the mass and charge of the black holes, we show that the presence of the aether field can also affect the size of the shadow. For the first type black hole, it is shown that the shadow size increases with the parameter $c_{13}$, while for the second type black hole, the shadow size still increases with c13 but decreases with the parameter $c_{14}$. With these properties of these aether parameters, we also discuss the observational constraints on these parameters by using the data of the first black hole image by the Event Horizon Telescope. In addition, we also explore the effect of the aether field on the the deflection angle of light and the time delay by using the Gauss-Bonnet theorem. It is shown that, for a specific combination $c_{123} = 0$, the deflection angle/time delay is slightly affected by the aether parameter $c_{13}$ at the leading order.

gr-qc↗

arXiv2019

Model Adaptation via Model Interpolation and Boosting for Web Search Ranking

This paper explores two classes of model adaptation methods for Web search ranking: Model Interpolation and error-driven learning approaches based on a boosting algorithm. The results show that model interpolation, though simple, achieves the best results on all the open test sets where the test data is very different from the training data. The tree-based boosting algorithm achieves the best performance on most of the closed test sets where the test data and the training data are similar, but its performance drops significantly on the open test sets due to the instability of trees. Several methods are explored to improve the robustness of the algorithm, with limited success.

cs.LG↗

arXiv2019

Sample Adaptive Multiple Kernel Learning for Failure Prediction of Railway Points

Railway points are among the key components of railway infrastructure. As a part of signal equipment, points control the routes of trains at railway junctions, having a significant impact on the reliability, capacity, and punctuality of rail transport. Traditionally, maintenance of points is based on a fixed time interval or raised after the equipment failures. Instead, it would be of great value if we could forecast points' failures and take action beforehand, minimising any negative effect. To date, most of the existing prediction methods are either lab-based or relying on specially installed sensors which makes them infeasible for large-scale implementation. Besides, they often use data from only one source. We, therefore, explore a new way that integrates multi-source data which are ready to hand to fulfil this task. We conducted our case study based on Sydney Trains rail network which is an extensive network of passenger and freight railways. Unfortunately, the real-world data are usually incomplete due to various reasons, e.g., faults in the database, operational errors or transmission faults. Besides, railway points differ in their locations, types and some other properties, which means it is hard to use a unified model to predict their failures. Aiming at this challenging task, we firstly constructed a dataset from multiple sources and selected key features with the help of domain experts. In this paper, we formulate our prediction task as a multiple kernel learning problem with missing kernels. We present a robust multiple kernel learning algorithm for predicting points failures. Our model takes into account the missing pattern of data as well as the inherent variance on different sets of railway points. Extensive experiments demonstrate the superiority of our algorithm compared with other state-of-the-art methods.

cs.LG↗

arXiv2019

Solo citations, duet citations, and prelude citations: New measures of the disruption of academic papers

It is important to measure the disruption of academic papers. According to the characteristics of three different kinds of citations, this paper borrows musical vocabulary and names them solo citations (SC), duet citations (DC), and prelude citations (PC) respectively. Studying how to measure the disruption of a published work effectively, this study analyzes nine indicators and suggests a general evaluation formula. Seven of the nine indicators are innovations introduced by this paper: SC, SC-DC, SC-PC, SC-DC-PC, (SC-DC)/(SC+DC), (SC-PC)/(SC+DC), and (SC-DC-PC)/(SC+DC), as is the general formula. These indices are discussed considering two cases: One case concerns the Citation Indexes for Science and the other concerns Co-citations. The results show that, compared with other indicators, four indicators (SC, SC-DC, SC/(SC+DC), and (SC-DC)/(SC+DC)) are logically and empirically reasonable. Future research may consider combining these indices, for example, using SC multiplied by SC/(SC+DC) or SC-DC multiplied by (SC-DC)/(SC+DC), to get final evaluation results that contain desirable characteristics of two types of indicators. Confirming which of the evaluation results from these indicators can best reflect the innovation of research papers requires much empirical analysis.

cs.DL↗

arXiv2019

Compare More Nuanced:Pairwise Alignment Bilinear Network For Few-shot Fine-grained Learning

The recognition ability of human beings is developed in a progressive way. Usually, children learn to discriminate various objects from coarse to fine-grained with limited supervision. Inspired by this learning process, we propose a simple yet effective model for the Few-Shot Fine-Grained (FSFG) recognition, which tries to tackle the challenging fine-grained recognition task using meta-learning. The proposed method, named Pairwise Alignment Bilinear Network (PABN), is an end-to-end deep neural network. Unlike traditional deep bilinear networks for fine-grained classification, which adopt the self-bilinear pooling to capture the subtle features of images, the proposed model uses a novel pairwise bilinear pooling to compare the nuanced differences between base images and query images for learning a deep distance metric. In order to match base image features with query image features, we design feature alignment losses before the proposed pairwise bilinear pooling. Experiment results on four fine-grained classification datasets and one generic few-shot dataset demonstrate that the proposed model outperforms both the state-ofthe-art few-shot fine-grained and general few-shot methods.

cs.CV↗

arXiv2019

Fast Registration for cross-source point clouds by using weak regional affinity and pixel-wise refinement

Many types of 3D acquisition sensors have emerged in recent years and point cloud has been widely used in many areas. Accurate and fast registration of cross-source 3D point clouds from different sensors is an emerged research problem in computer vision. This problem is extremely challenging because cross-source point clouds contain a mixture of various variances, such as density, partial overlap, large noise and outliers, viewpoint changing. In this paper, an algorithm is proposed to align cross-source point clouds with both high accuracy and high efficiency. There are two main contributions: firstly, two components, the weak region affinity and pixel-wise refinement, are proposed to maintain the global and local information of 3D point clouds. Then, these two components are integrated into an iterative tensor-based registration algorithm to solve the cross-source point cloud registration problem. We conduct experiments on synthetic cross-source benchmark dataset and real cross-source datasets. Comparison with six state-of-the-art methods, the proposed method obtains both higher efficiency and accuracy.

cs.CV↗

arXiv2018

Higher-dimension Tensor Completion via Low-rank Tensor Ring Decomposition

The problem of incomplete data is common in signal processing and machine learning. Tensor completion algorithms aim to recover the incomplete data from its partially observed entries. In this paper, taking advantages of high compressibility and flexibility of recently proposed tensor ring (TR) decomposition, we propose a new tensor completion approach named tensor ring weighted optimization (TR-WOPT). It finds the latent factors of the incomplete tensor by gradient descent algorithm, then the latent factors are employed to predict the missing entries of the tensor. We conduct various tensor completion experiments on synthetic data and real-world data. The simulation results show that TR-WOPT performs well in various high-dimension tensors. Furthermore, image completion results show that our proposed algorithm outperforms the state-of-the-art algorithms in many situations. Especially when the missing rate of the test images is high (e.g., over 0.9), the performance of our TR-WOPT is significantly better than the compared algorithms.

math.NA↗

arXiv2018

Non-adiabatic Evolution of Primordial Perturbations and non-Gaussinity in Hybrid Approach of Loop Quantum Cosmology

While loop quantum cosmology (LQC) predicts a robust quantum bounce of the background evolution of a Friedmann-Robertson-Walker (FRW) spacetime prior to the standard slow-roll inflation, whereby the big bang singularity is resolved, there are several different quantization procedures to cosmological perturbations, for instance, {\em the deformed algebra, dressed metric, and hybrid quantizations}. This paper devotes to study the quantum bounce effects of primordial perturbations in the hybrid approach. The main discrepancy of this approach is the effective positive mass at the quantum bounce for the evolution of the background that is dominated by the kinetic energy of the inflaton field at the bounce, while this mass is always nonpositive in the dressed metric approach. It is this positivity of the effective mass that violates the adiabatic evolution of primordial perturbations at the initial moments of the quantum bounce. With the assumption that the evolution of the background is dominated by the kinetic energy of the inflaton at the bounce, we find that the effective potentials for both scalar and tensor perturbations can be well approximately described by a Pöschl-Teller (PT) potential, which allows us to find analytical solutions of perturbations, and from these analytical expressions we are able to study the non-adiabatic evolution of primordial perturbations in details. In particular, we derive their quantum bounce effects and investigate their observational constraints. In addition, the impacts of quantum bounce effects on the non-Gaussinity and their implication on the explanations of observed power asymmetry in CMB have also been explored.

gr-qc↗