arXiv subjects

Yi Wang

Publications and source records attributed to Yi Wang.

At least 829 records · Page 46Linked to original sources

arXiv2020

Synthesis and Edition of Ultrasound Images via Sketch Guided Progressive Growing GANs

Ultrasound (US) is widely accepted in clinic for anatomical structure inspection. However, lacking in resources to practice US scan, novices often struggle to learn the operation skills. Also, in the deep learning era, automated US image analysis is limited by the lack of annotated samples. Efficiently synthesizing realistic, editable and high resolution US images can solve the problems. The task is challenging and previous methods can only partially complete it. In this paper, we devise a new framework for US image synthesis. Particularly, we firstly adopt a sketch generative adversarial networks (Sgan) to introduce background sketch upon object mask in a conditioned generative adversarial network. With enriched sketch cues, Sgan can generate realistic US images with editable and fine-grained structure details. Although effective, Sgan is hard to generate high resolution US images. To achieve this, we further implant the Sgan into a progressive growing scheme (PGSgan). By smoothly growing both generator and discriminator, PGSgan can gradually synthesize US images from low to high resolution. By synthesizing ovary and follicle US images, our extensive perceptual evaluation, user study and segmentation results prove the promising efficacy and efficiency of the proposed PGSgan.

eess.IV↗

arXiv2020

On the largest eigenvalue of a mixed graph with partial orientation

Let $G$ be a connected graph and let $T$ be a spanning tree of $G$. A partial orientation $σ$ of $G$ respect to $T$ is an orientation of the edges of $G$ except those edges of $T$, the resulting graph associated with which is denoted by $G_T^σ$. In this paper we prove that there exists a partial orientation $σ$ of $G$ respect to $T$ such that the largest eigenvalue of the Hermitian adjacency matrix of $G_T^σ$ is at most the largest absolute value of the roots of the matching polynomial of $G$.

math.CO↗

arXiv2020

VCNet: A Robust Approach to Blind Image Inpainting

Blind inpainting is a task to automatically complete visual contents without specifying masks for missing areas in an image. Previous works assume missing region patterns are known, limiting its application scope. In this paper, we relax the assumption by defining a new blind inpainting setting, making training a blind inpainting neural system robust against various unknown missing region patterns. Specifically, we propose a two-stage visual consistency network (VCN), meant to estimate where to fill (via masks) and generate what to fill. In this procedure, the unavoidable potential mask prediction errors lead to severe artifacts in the subsequent repairing. To address it, our VCN predicts semantically inconsistent regions first, making mask prediction more tractable. Then it repairs these estimated missing regions using a new spatial normalization, enabling VCN to be robust to the mask prediction errors. In this way, semantically convincing and visually compelling content is thus generated. Extensive experiments are conducted, showing our method is effective and robust in blind image inpainting. And our VCN allows for a wide spectrum of applications.

cs.CV↗

arXiv2020

Quantum States of Higher-order Whispering gallery modes in a Silicon Micro-disk Resonator

The quantum states of light in an integrated photonics platform provide an important resource for quantum information processing and takes advantage of the scalability and practicality of silicon photonics. Integrated resonators have been well explored in classical and quantum optics. However, to encode multiple information through integrated quantum optics requires broader utilization of the available degrees of freedom on a chip. Here, we studied the quantum interference between photon pairs of the same higher order whispering gallery modes populated by spontaneous four-wave mixing in an integrated silicon micro-disk resonator. The quantum interference between the photon pairs of the first two quasi-TE0 and quasi-TE1 radial modes was measured to be Vnet ~ 98 + 0.8 % and Vnet ~ 94 + 2.6 %, respectively. The results are promising for achieving higher-dimensional quantum states using the higher-order radial modes of a micro-disk resonator coupled with an integrated waveguide.

physics.optics↗

arXiv2020

Minimal non-odd-transversal hypergraphs and minimal non-odd-bipartite hypergraphs

Among all uniform hypergraphs with even uniformity, the odd-transversal or odd-bipartite hypergraphs are more close to bipartite simple graphs from the viewpoint of both structure and spectrum. A hypergraph is called minimal non-odd-transversal if it is non-odd-transversal but deleting any edge results in an odd-transversal hypergraph. In this paper we give an equivalent characterization of the minimal non-odd-transversal hypergraphs by the degrees and the rank of its incidence matrix over $\mathbb{Z}_2$. If a minimal non-odd-transversal hypergraph is uniform, then it has even uniformity, and hence is minimal non-odd-bipartite. We characterize $2$-regular uniform minimal non-odd-bipartite hypergraphs, and give some examples of $d$-regular uniform hypergraphs which are minimal non-odd-bipartite. Finally we give upper bounds for the least H-eigenvalue of the adjacency tensor of minimal non-odd-bipartite hypergraphs.

math.CO↗

arXiv2020

RSANet: Recurrent Slice-wise Attention Network for Multiple Sclerosis Lesion Segmentation

Brain lesion volume measured on T2 weighted MRI images is a clinically important disease marker in multiple sclerosis (MS). Manual delineation of MS lesions is a time-consuming and highly operator-dependent task, which is influenced by lesion size, shape and conspicuity. Recently, automated lesion segmentation algorithms based on deep neural networks have been developed with promising results. In this paper, we propose a novel recurrent slice-wise attention network (RSANet), which models 3D MRI images as sequences of slices and captures long-range dependencies through a recurrent manner to utilize contextual information of MS lesions. Experiments on a dataset with 43 patients show that the proposed method outperforms the state-of-the-art approaches. Our implementation is available online at https://github.com/tinymilky/RSANet.

eess.IV↗

arXiv2020

Privacy-Preserving Distributed Clustering for Electrical Load Profiling

Electrical load profiling supports retailers and distribution network operators in having a better understanding of the consumption behavior of consumers. However, traditional clustering methods for load profiling are centralized and require access to all the smart meter data, thus causing privacy issues for consumers and retailers. To tackle this issue, we propose a privacy-preserving distributed clustering framework for load profiling by developing a privacy-preserving accelerated average consensus (PP-AAC) algorithm with proven convergence. Using the proposed framework, we modify several commonly used clustering methods, including k-means, fuzzy C-means, and Gaussian mixture model, to provide privacy-preserving distributed clustering methods. In this way, load profiling can be performed only by local calculations and information sharing between neighboring data owners without sacrificing privacy. Meanwhile, compared to traditional centralized clustering methods, the computational time consumed by each data owner is significantly reduced. The privacy and complexity of the proposed privacy-preserving distributed clustering framework are analyzed. The correctness, efficiency, effectiveness, and privacy-preserving feature of the proposed framework and the proposed PP-AAC algorithm are verified using a real-world Irish residential dataset.

eess.SY↗

arXiv2020

The least H-eigenvalue of adjacency tensor of hypergraphs with cut vertices

Let $G$ be a connected hypergraph with even uniformity, which contains cut vertices. Then $G$ is the coalescence of two nontrivial connected sub-hypergraphs (called branches) at a cut vertex. Let $\mathcal{A}(G)$ be the adjacency tensor of $G$. The least H-eigenvalue of $\mathcal{A}(G)$ refers to the least real eigenvalue of $\mathcal{A}(G)$ associated with a real eigenvector. In this paper we obtain a perturbation result on the least H-eigenvalue of $\mathcal{A}(G)$ when a branch of $G$ attached at one vertex is relocated to another vertex, and characterize the unique hypergraph whose least H-eigenvalue attains the minimum among all hypergraphs in a certain class of hypergraphs which contain a fixed connected hypergraph.

math.CO↗

arXiv2020

Data-driven MHD Simulation of the Formation and Initiation of a Large-scale Pre-flare Magnetic Flux Rope in Solar Active Region 12371

Solar eruptions are the most powerful drivers of space weather. To understand their cause and nature, it is crucial to know how the coronal magnetic field evolves before eruption. Here we study the formation process of a relatively large-scale magnetic flux rope (MFR) in active region NOAA~12371 that erupts with a major flare and coronal mass ejection on 2015 June 21. A data-driven numerical magnetohydrodynamic model is employed to simulate three-dimensional coronal magnetic field evolution of one-day duration before the eruption. Comparison between the observed features and our modeled magnetic field discloses how the pre-eruption MFR forms. Initially, the magnetic field lines were weakly twisted as being simple sheared arcades. Then a long MFR was formed along the polarity inversion line due to the complex photospheric motion, which is mainly shearing rather than twisting. The presence of the MFR is evidenced by a coherent set of magnetic field lines with twist number above unity. Below the MFR a current sheet is shown in the model, suggesting that tether-cutting reconnection plays a key role in the MFR formation. The MFR's flux grows as more and more field lines are twisted due to continuous injection of magnetic helicity by the photospheric motions. Meanwhile, the height of the MFR's axis increases monotonely from its formation. By an analysis of the decay index of its overlying field, we suggest that it is because the MFR runs into the torus instability regime and becomes unstable that finally triggers the eruption.

astro-ph.SR↗

arXiv2020

Cosmological Signatures of Superheavy Dark Matter

We discuss two possible scenarios, namely the curvaton mechanism and the dark matter density modulation, where non-Gaussianity signals of superheavy dark matter produced by gravity can be enhanced and observed. In both scenarios, superheavy dark matter couples to an additional light field as a mediator. In the case of derivative coupling, the resulting non-Gaussianities induced by the light field can be large, which can provide inflationary evidences for these superheavy dark matter scenarios.

hep-ph↗

arXiv2020

Controlled release of entrapped nanoparticles from thermoresponsive hydrogels with tunable network characteristics

Thermoresponsive hydrogels have been studied intensively for creating smart drug carriers and controlled drug delivery. Understanding the drug release kinetics and corresponding transport mechanisms of nanoparticles (NPs) in a thermoresponsive hydrogel network is the key to the successful design of a smart drug delivery system. We construct a mesoscopic model of rigid NPs entrapped in a hydrogel network in an aqueous solution, where the hydrogel network is formed by cross-linked semiflexible polymers of PNIPAM. By varying the environmental temperature crossing the lower critical solution temperature of PNIPAM we can significantly change the hydrogel network characteristics. We systematically investigate how the matrix porosity and the nanoparticle size affect the NPs' transport kinetics at different temperatures. Quantitative results on the mean-squared displacement and the van Hove displacement distributions of NPs show that all NPs entrapped in the smart hydrogels undergo subdiffusion at both low and high temperatures. For a coil state, the subdiffusive exponent and the diffusion coefficient of NPs increase due to the increased kinetic energy and the decreased confinement on NPs, while the transport of NPs in the hydrogels can be also enhanced by decreasing the matrix porosity and NPs' size. However, when the solution temperature is increased above the critical temperature, the hydrogel network collapses following the coil-to-globule transition, with the NPs tightly trapped in some local regions inside the hydrogels. Consequently, the NP diffusion coefficient can be reduced by two orders of magnitude, or the diffusion processes can even be completely stopped. These findings provide new insights for designing controlled drug release from stimuli-responsive hydrogels, including autonomously switch on/off drug release to respond to the changes of the local environment.

cond-mat.soft↗

arXiv2020

SQLFlow: A Bridge between SQL and Machine Learning

Industrial AI systems are mostly end-to-end machine learning (ML) workflows. A typical recommendation or business intelligence system includes many online micro-services and offline jobs. We describe SQLFlow for developing such workflows efficiently in SQL. SQL enables developers to write short programs focusing on the purpose (what) and ignoring the procedure (how). Previous database systems extended their SQL dialect to support ML. SQLFlow (https://sqlflow.org/sqlflow ) takes another strategy to work as a bridge over various database systems, including MySQL, Apache Hive, and Alibaba MaxCompute, and ML engines like TensorFlow, XGBoost, and scikit-learn. We extended SQL syntax carefully to make the extension working with various SQL dialects. We implement the extension by inventing a collaborative parsing algorithm. SQLFlow is efficient and expressive to a wide variety of ML techniques -- supervised and unsupervised learning; deep networks and tree models; visual model explanation in addition to training and prediction; data processing and feature extraction in addition to ML. SQLFlow compiles a SQL program into a Kubernetes-native workflow for fault-tolerable execution and on-cloud deployment. Current industrial users include Ant Financial, DiDi, and Alibaba Group.

cs.DB↗

arXiv2020

Continuous Null-Point Magnetic Reconnection Builds Up a Torus Unstable Magnetic Flux Rope Triggering the X9.3 Flare in Solar Active Region~12673

Two X-class solar flares occurred on 2017 September 6 from active region NOAA 12673: the first one is a confined X2.2 flare, and it is followed only $\sim 3$ hours later by the second one, which is the strongest flare in solar cycle 24, reaching X9.3 class and accompanied with a coronal mass ejection. Why these two X-class flares occurred in the same position with similar magnetic configurations, but one is eruptive while the other is not? Here we track the coronal magnetic field evolution via nonlinear force-free field extrapolations from a time sequence of vector magnetograms with high cadence. A detailed analysis of the magnetic field shows that a magnetic flux rope (MFR) forms and grows gradually before the first flare, and shortly afterwards, the MFR's growth is significantly enhanced with a much faster rise in height, from far below the threshold of torus instability to above it, while the magnetic twist only increases mildly. Combining EUV observations and the magnetic field extrapolation, we found that overlying the MFR is a null-point magnetic topology, where recurrent brightening is seen after the first flare. We thus suggest a scenario to interpret the occurrence of the two flares. The first flare occurred since the MFR reached a high enough height to activate the null point, and its continuous expansion forces the null-point reconnection recurrently. Such reconnection weakens the overlying field, allowing the MFR to rise faster, which eventually crosses the threshold of torus instability and triggers the second, eruptive flare.

astro-ph.SR↗

arXiv2020

Cosmological Collider Signatures of Massive Vectors from Non-Gaussian Gravitational Waves

The cosmological collider provides a model-independent probe of particle physics during inflation. We extend the study of cosmological collider physics to much smaller scales through gravitational wave (GW) probes. With a Chern-Simons interaction, a massive vector field can obtain a chemical potential and its particle production can cause significant non-Gaussian GW signals. We calculate the mass and spin dependences of the induced GW 3-point correlation function in the squeezed limit, and estimate its amplitude. Such signals may be detectable in the current and upcoming GW interferometer experiments.

astro-ph.CO↗

arXiv2020

A3: An Automatic Topology-Aware Malfunction Detection and Fixation System in Data Center Networks

Link failures and cable miswirings are not uncommon in building data center networks, which prevents the existing automatic address configuration methods from functioning correctly. However, accurately detecting such malfunctions is not an easy task because there could be no observable node degree changes. Fixing or correcting such malfunctions is even harder as almost no work can provide accurate fixation suggestions now. To solve the problems, we design and implement A3, an automatic topology-aware malfunction detection and fixation system. A3 innovatively formulates the problem of finding minimal fixation to the problem of computing minimum graph difference (NP-hard) and solves it in O(k^6) and O(k^3) for any less than k/2 and k/4 undirected link malfunctions for FatTree, respectively. Our evaluation demonstrates that for less than k/2 undirected link malfunctions, A3 is 100% accurate for malfunction detection and provides the minimum fixation result. For greater or equal to k/2 undirected link malfunctions, A3 still has accuracy of about 100% and provides the near optimal fixation result.

cs.NI↗

arXiv2019

A gigaparsec-scale local void and the Hubble tension

We explore the possibility of using a gigaparsec-scale local void to reconcile the Hubble tension. Such a gigaparsec-scale void can be produced by multi-stream inflation where different parts of the observable universe follow different inflationary trajectories. The impact of such a void for cosmological observations is studied, especially those involving supernovae, Baryon Acoustic Oscillations (BAO) and the kinetic Sunyaev-Zel'dovich (kSZ) effect. As a benchmark model, a 1.7Gpc scale with boundary width 0.7Gpc and density contrast -0.14 may ease the Hubble tension.

astro-ph.CO↗

arXiv2019

Turaev Hyperbolicity of Classical and Virtual Knots

By work of W. Thurston, knots and links in the 3-sphere are known to either be torus links, or to contain an essential torus in their complement, or to be hyperbolic, in which case a unique hyperbolic volume can be calculated for their complement. We employ a construction of Turaev to associate a family of hyperbolic 3-manifolds of finite volume to any classical or virtual link, even if non-hyperbolic. These are in turn used to define the Turaev volume of a link, which is the minimal volume among all the hyperbolic 3-manifolds associated via this Turaev construction. In the case of a classical link, we can also define the classical Turaev volume, which is the minimal volume among all the hyperbolic 3-manifolds associated via this Turaev construction for the classical projections only. We then investigate these new invariants.

math.GT↗

arXiv2019

Analytic properties of sextet polynomials of hexagonal systems

In this paper we investigate analytic properties of sextet polynomials of hexagonal systems. For the pyrene chains, we show that zeros of the sextet polynomials $P_n(x)$ are real, located in the open interval $(-3-2\sqrt{2},-3+2\sqrt{2})$ and dense in the corresponding closed interval. We also show that coefficients of $P_n(x)$ are symmetric, unimodal, log-concave, and asymptotically normal. For general hexagonal systems, we show that real zeros of all sextet polynomials are dense in the interval $(-\infty,0]$, and conjecture that every sextet polynomial has log-concave coefficients.

math.CO↗