arXiv subjects

Kun Wang

Publications and source records attributed to Kun Wang.

At least 469 records · Page 26Linked to original sources

arXiv2020

Permutation Enhances Classical Communication Assisted by Entangled States

We give a capacity formula for the classical communication over a noisy quantum channel, when local operations and global permutations allowed in the encoding and bipartite states preshared between the sender and the receiver. The two endpoints of this formula are the Holevo capacity (without entanglement assistance) and the entanglement-assisted capacity (with unlimited entanglement assistance). What's more, we show that the capacity satisfies the strong converse property and thus the formula serves as a sharp dividing line between achievable and unachievable rates of communication. We prove that the difference between the assisted capacity and the Holevo capacity is upper bounded by the discord of formation of the preshared state. As examples, we derive analytically the classical capacity of various quantum channels of interests. Our result witnesses the power of random permutation in classical communication, whenever entanglement assistance is available.

quant-ph↗

arXiv2019

CompareNet: Anatomical Segmentation Network with Deep Non-local Label Fusion

Label propagation is a popular technique for anatomical segmentation. In this work, we propose a novel deep framework for label propagation based on non-local label fusion. Our framework, named CompareNet, incorporates subnets for both extracting discriminating features, and learning the similarity measure, which lead to accurate segmentation. We also introduce the voxel-wise classification as an unary potential to the label fusion function, for alleviating the search failure issue of the existing non-local fusion strategies. Moreover, CompareNet is end-to-end trainable, and all the parameters are learnt together for the optimal performance. By evaluating CompareNet on two public datasets IBSRv2 and MICCAI 2012 for brain segmentation, we show it outperforms state-of-the-art methods in accuracy, while being robust to pathologies.

cs.CV↗

arXiv2019

Calibration of X-ray telescope prototypes at PANTER

We report a ground X-ray calibration of two X-ray telescope prototypes at the PANTER X-ray Test Facility, of the Max-Planck-Institute for Extraterrestrial Physics, in Neuried, Germany. The X-ray telescope prototypes were developed by the Institute of Precision Optical Engineering (IPOE) of Tongji University, in a conical Wolter-I configuration, using thermal glass slumping technology. Prototype #1 with 3 layers and Prototype #2 with 21 layers were tested to assess the prototypes' on-axis imaging performance. The measurement of Prototype #1 indicates a Half Power Diameter (HPD) of 82" at 1.49 keV. As for Prototype #2, we performed more comprehensive measurements of on-axis angular resolution and effective area at several energies ranging from 0.5-10 keV. The HPD and effective area are 111" and 39 cm^2 at 1.49 keV, respectively, at which energy the on-axis performance of the prototypes is our greatest concern.

astro-ph.IM↗

arXiv2019

On the Decomposition Theorems for C*-algebras

Elliott dimension drop interval algebra is an important class among all $C^*$-algebras in the classification theory. Especially, they are building stones of $\mathcal{AHD}$ algebra and the latter contains all $AH$ algebras with the ideal property of no dimension growth. In this paper, we will show two decomposition theorems related to the Elliott dimension drop interval algebra. Our results are key steps in classifying all $AH$ algebras with the ideal property of no dimension growth.

math.OA↗

arXiv2019

Code-Switching for Enhancing NMT with Pre-Specified Translation

Leveraging user-provided translation to constrain NMT has practical significance. Existing methods can be classified into two main categories, namely the use of placeholder tags for lexicon words and the use of hard constraints during decoding. Both methods can hurt translation fidelity for various reasons. We investigate a data augmentation method, making code-switched training data by replacing source phrases with their target translations. Our method does not change the MNT model or decoding algorithm, allowing the model to learn lexicon translations by copying source-side target words. Extensive experiments show that our method achieves consistent improvements over existing approaches, improving translation of constrained words without hurting unconstrained words.

cs.CL↗

arXiv2019

Reconstruction-Aware Imaging System Ranking by use of a Sparsity-Driven Numerical Observer Enabled by Variational Bayesian Inference

It is widely accepted that optimization of imaging system performance should be guided by task-based measures of image quality (IQ). It has been advocated that imaging hardware or data-acquisition designs should be optimized by use of an ideal observer (IO) that exploits full statistical knowledge of the measurement noise and class of objects to be imaged, without consideration of the reconstruction method. In practice, accurate and tractable models of the complete object statistics are often difficult to determine. Moreover, in imaging systems that employ compressive sensing concepts, imaging hardware and sparse image reconstruction are innately coupled technologies. In this work, a sparsity-driven observer (SDO) that can be employed to optimize hardware by use of a stochastic object model describing object sparsity is described and investigated. The SDO and sparse reconstruction method can therefore be "matched" in the sense that they both utilize the same statistical information regarding the class of objects to be imaged. To efficiently compute the SDO test statistic, computational tools developed recently for variational Bayesian inference with sparse linear models are adopted. The use of the SDO to rank data-acquisition designs in a stylized example as motivated by magnetic resonance imaging (MRI) is demonstrated. This study reveals that the SDO can produce rankings that are consistent with visual assessments of the reconstructed images but different from those produced by use of the traditionally employed Hotelling observer (HO).

cs.CV↗

arXiv2019

Time-sync Video Tag Extraction Using Semantic Association Graph

Time-sync comments reveal a new way of extracting the online video tags. However, such time-sync comments have lots of noises due to users' diverse comments, introducing great challenges for accurate and fast video tag extractions. In this paper, we propose an unsupervised video tag extraction algorithm named Semantic Weight-Inverse Document Frequency (SW-IDF). Specifically, we first generate corresponding semantic association graph (SAG) using semantic similarities and timestamps of the time-sync comments. Second, we propose two graph cluster algorithms, i.e., dialogue-based algorithm and topic center-based algorithm, to deal with the videos with different density of comments. Third, we design a graph iteration algorithm to assign the weight to each comment based on the degrees of the clustered subgraphs, which can differentiate the meaningful comments from the noises. Finally, we gain the weight of each word by combining Semantic Weight (SW) and Inverse Document Frequency (IDF). In this way, the video tags are extracted automatically in an unsupervised way. Extensive experiments have shown that SW-IDF (dialogue-based algorithm) achieves 0.4210 F1-score and 0.4932 MAP (Mean Average Precision) in high-density comments, 0.4267 F1-score and 0.3623 MAP in low-density comments; while SW-IDF (topic center-based algorithm) achieves 0.4444 F1-score and 0.5122 MAP in high-density comments, 0.4207 F1-score and 0.3522 MAP in low-density comments. It has a better performance than the state-of-the-art unsupervised algorithms in both F1-score and MAP.

cs.IR↗

arXiv2019

Extended main-sequence turnoffs in the double cluster $h$ and $χ$ Persei: The complex role of stellar rotation

Using {\sl Gaia} Data Release 2 photometry, we report the detection of extended main-sequence turnoff (eMSTO) regions in the color--magnitude diagrams (CMDs) of the $\sim 14$ Myr-old double clusters $h$ and $χ$ Persei (NGC 869 and NGC 884). We find that stars with masses below $\sim$1.3 $M_{\odot}$ in both $h$ and $χ$ Persei populate narrow main sequences (MSs), while more massive stars define the eMSTO, closely mimicking observations of young Galactic and Magellanic Cloud clusters (with ages older than $\sim$30 Myr). Previous studies based on clusters older than $\sim$30 Myr find that rapidly rotating MS stars are redder than slow rotators of similar luminosity, suggesting that stellar rotation may be the main driver of the eMSTO. By combining photometry and projected rotational velocities from the literature of stars in $h$ and $χ$ Persei, we find no obvious relation between the rotational velocities and colors of non-emission-line eMSTO stars, in contrast with what is observed in older clusters. Similarly to what is observed in Magellanic Cloud clusters, most of the extremely rapidly rotating stars, identified by their strong H$α$ emission lines, are located in the red part of the eMSTOs. This indicates that stellar rotation plays a role in the color and magnitude distribution of MSTO stars. By comparing the observations with simulated CMDs, we find that a simple population composed of coeval stars that span a wide range of rotation rates is unable to reproduce the color spread of the clusters' MSs. We suggest that variable stars, binary interactions, and stellar rotation affect the eMSTO morphology of these very young clusters.

astro-ph.SR↗

arXiv2019

Attention-based Convolutional Neural Network for Weakly Labeled Human Activities Recognition with Wearable Sensors

Unlike images or videos data which can be easily labeled by human being, sensor data annotation is a time-consuming process. However, traditional methods of human activity recognition require a large amount of such strictly labeled data for training classifiers. In this paper, we present an attention-based convolutional neural network for human recognition from weakly labeled data. The proposed attention model can focus on labeled activity among a long sequence of sensor data, and while filter out a large amount of background noise signals. In experiment on the weakly labeled dataset, we show that our attention model outperforms classical deep learning methods in accuracy. Besides, we determine the specific locations of the labeled activity in a long sequence of weakly labeled data by converting the compatibility score which is generated from attention model to compatibility density. Our method greatly facilitates the process of sensor data annotation, and makes data collection more easy.

cs.LG↗

arXiv2019

A cooperative game for automated learning of elasto-plasticity knowledge graphs and models with AI-guided experimentation

We introduce a multi-agent meta-modeling game to generate data, knowledge, and models that make predictions on constitutive responses of elasto-plastic materials. We introduce a new concept from graph theory where a modeler agent is tasked with evaluating all the modeling options recast as a directed multigraph and find the optimal path that links the source of the directed graph (e.g. strain history) to the target (e.g. stress) measured by an objective function. Meanwhile, the data agent, which is tasked with generating data from real or virtual experiments (e.g. molecular dynamics, discrete element simulations), interacts with the modeling agent sequentially and uses reinforcement learning to design new experiments to optimize the prediction capacity. Consequently, this treatment enables us to emulate an idealized scientific collaboration as selections of the optimal choices in a decision tree search done automatically via deep reinforcement learning.

cs.LG↗

arXiv2019

Optimal Verification of Two-Qubit Pure States

In a recent work [Phys. Rev. Lett. 120, 170502 (2018)], Pallister et al. proposed an optimal strategy to verify non-maximally entangled two-qubit pure states under the constraint that the accessible measurements being locally projective and non-adaptive. Their nice result leads naturally to the question: What is the optimal strategy among general LOCC measurements? In this Letter, we answer this problem completely for two-qubit pure states. To be specific, we give the optimal strategy for each of the following available classes of measurements: (i) local operations and one-way classical communication (one-way LOCC) measurements; (ii) local operations and two-way classical communication (two-way LOCC) measurements; and (iii) separable measurements. Surprisingly, our results reveal that for the two-qubit pure state verification problem, two-way LOCC measurements remarkably outperforms one-way LOCC measurements and has the same power as the separable measurements.

quant-ph↗

arXiv2019

Frequent flare events on the short-period M-type eclipsing binary BX Tri

We present long-term, multi-color photometric monitoring and spectroscopic observations of the short-period M-type eclipsing binary BX Tri. Six flare events were recorded in 4 years from 2014 to 2017. Three of them were detected on one night within an orbital cycle. The strongest one was identified on December 23, 2014. With the amplitudes of $Δ$B=0.48 mag, $Δ$V=0.28 mag, $Δ$R=0.10 mag, $Δ$I=0.02 mag, the total energy due to the flare event was measured to be 4.08 ($\pm$0.24) $\times$$10^{34}$ erg, exceeding the superflare level ($10^{34}$). Based on the observations, the evolutionary status of the binary system as well as the long-term orbital period changes were analyzed. It reveals that BX Tri is probably a semi-detached system with the primary component filling its Roche lobe. The extremely high occurrence of flare events of the binary could be related to the rapid mass transfer between components.

astro-ph.SR↗

arXiv2018

Negative differential conductance effect and electrical anisotropy of 2D ZrB2 monolayers

Two-dimensional (2D) metal-diboride ZrB2 monolayers was predicted theoretically as a stable new electronic material [A. Lopez-Bezanilla, Phys. Rev. Mater., 2018, 2, 011002 (R)]. Here, we investigate its electronic transport properties along the zigzag (z-ZrB2) and armchair (a-ZrB2) directions, using the density functional theory and non-equilibrium Green's function methods. Under low biases, the 2D ZrB2 shows a similar electrical transport along zigzag and armchair directions as electric current propagates mostly via the metallic Zr-Zr bonds. However, it shows an electrical anistropy under high biases, and its I-V curves along zigzag and armchair directions diverge as the bias voltage is higher than 1.4 V, as more directional B-B transmission channels are opened. Importantly, both z-ZrB2 and a-ZrB2 show a pronounced negative differential conductance (NDC) effect and hence they can be promising for the use in NDC-based nanodevices.

cond-mat.mes-hall↗

arXiv2018

Solving the muon g-2 anomaly in CMSSM extension with non-universal gaugino masses

We propose to generate non-universal gaugino masses in SU(5) GUT with the generalized Planck-scale mediation SUSY breaking mechanism, in which the non-universality arises from proper wavefunction normalization with lowest component VEVs of various high dimensional representations of the Higgs fields of SU(5) and an unique F-term VEV by the singlet. Different predictions on gaugino mass ratios with respect to widely studied scenarios are given. The gluino-SUGRA-like scenario, where gluinos are much heavier than winos, bino and universal scalar masses, can be easily realized with appropriate combinations of such high-representation Higgs fields. With 6 GUT-scale parameters in our scenario, we can solve elegantly the tension between mSUGRA and the present experimental results, including the muon g-2, the dark matter (DM) relic density and the direct sparticle search bounds from the LHC. Taking into account the current constraints in our numerical scan, we have the following observations: (i) The large-$\tanβ$ samples with a moderate $M_3$ a small $|A_0/M_3|$ and a small $m_A$ are favored to generate a 125 GeV SM-like Higgs and predict a large muon g-2, while the stop mass and μparameter, mainly determined by $|M_3|(\gg M_0,|M_1|,|M_2|)$, can be about 6 TeV; (ii) The moderate-$\tanβ$(35~40) samples with a negative $M_3$ can have a light smuon but a heavy stau, which predict a large muon g-2 but a small $Br(B_s\toμ^+μ^-)$; (iii) To obtain the right DM relic density, the annihilation mechanisms should be stau exchange, stau coannihilation, chargino coannihilation, slepton annihilation and the combination of two or three of them; (iv) To obtain the right DM relic density, the spin-independent DM-nucleon cross section is typically much smaller than the present limits of XENON1T 2018 and also an order of magnitude lower than the future detection sensitivity of LZ and XENONnT.

hep-ph↗

arXiv2018

A Light Scalar in the Minimal Dilaton Model in Light of LHC Constraints

Whether an additional light scalar exists is an interesting topic beyond the Standard Model (SM), while nowadays we do not know exactly physics beyond the SM in the low mass region, e.g., the Atlas and CMS collaborations get inconsistent results at around 95 GeV in searching for light resonances in diphoton channel. Considering these, we study a light scalar in the Minimal Dilaton Model (MDM). Under the theoretical and latest experimental constraints, we sort the surviving samples into two scenarios according to the diphoton rate of the light scalar: the large-diphoton scenario (with $/SM\gtrsim0.2$) and the small-diphoton scenario (with $/SM\lesssim0.2$), which are favored by CMS and Atlas results respectively. We compare the two scenarios, check the characteristics in model parameters, scalar couplings, production and decay, and consider further distinguishing them at colliders. Finally, we get the following conclusions for the two scenarios: (i) The formal usually has small Higgs-dilaton mixing angle ($|\sinθ_S|\lesssim0.2$) and small dilaton vacuum expectation value (VEV) $f$ ($0.5\lesssimη\equiv v/f\lesssim1$), and the later usually has large mixing ($|\sinθ_S|\gtrsim0.4$) or large VEV ($η\equiv v/f\lesssim0.3$). (ii) The former usually predicts small $sγγ$ coupling ($|C_{sγγ}/SM|\lesssim0.3$) and large $sgg$ coupling ($0.6\lesssim|C_{sgg}/SM|\lesssim1.2$), while the later usually predicts small $sgg$ coupling ($|C_{sgg}/SM|\lesssim0.5$). (iii) The former can interpret the small diphoton excess by CMS at its central value, when $m_s\simeq$95GeV, $η\simeq0.6$ and $|\sinθ_S|\simeq0$. (iv) The former usually predicts a negative correlation between Higgs couplings $|C_{hγγ}/SM|$ and $|C_{hgg}/SM|$, while the later usually predicts the two couplings both smaller than 1, or $|C_{hγγ}/SM|\lesssim0.9 \lesssim|C_{hgg}/SM|$.

hep-ph↗

arXiv2018

The semi-constrained NMSSM in light of muon g-2, LHC, and dark matter constraints

The semi-constrained NMSSM (scNMSSM) extends the MSSM by a singlet field, and requires unification of the soft SUSY breaking terms in the squark and slepton sectors, while it allows that in the Higgs sector to be different. We try to interpret the muon g-2 in the scNMSSM, under the constraints of 125 GeV Higgs data, B physics, searches for low and high mass resonances, searches for SUSY particles at the LHC, dark matter relic density by WMAP/Planck, and direct searches for dark matter by LUX, XENON1T, and PandaX-II. We find that under the above constraints, the scNMSSM can still (i) satisfy muon g-2 at 1$σ$ level, with a light muon sneutrino and light chargino; (ii) predict a highly-singlet-dominated 95~GeV Higgs, with a diphoton rate as hinted at by CMS data, because of a light higgsino-like chargino and moderate $λ$; (iii) get low fine tuning from the GUT scale with small $μ_{\rm eff},\, M_0,\, M_{1/2},\, {\rm and}\, A_0$, with a lighter stop mass which can be as low as about 500 GeV, which can be further checked in future studies with search results from the 13~TeV LHC; (iv) have the lightest neutralino be singlino-dominated or higgsino-dominated, while the bino and wino are heavier because of high gluino bounds at the LHC and universal gaugino conditions at the GUT scale; (v) satisfy all the above constraints, although it is not easy for the lightest neutralino, as the only dark matter candidate, to get enough relic density. Several ways to increase relic density are discussed.

hep-ph↗

arXiv2018

Robustness of link prediction under network attacks

While link prediction in networks has been a hot topic over the years, its robustness has not been well discussed in literature. In this paper, we study the robustness of some mainstream link prediction methods under various kinds of network attack strategies, including the random attack (RDA), centrality based attacks (CA), similarity based attacks (SA), and simulated annealing based attack (SAA). Through the variation of precision, a typical evaluation index of link prediction, we find that for the SA and SAA, a small fraction of link removals can significantly reduce the performance of link prediction. In general, the SAA has the highest attack efficiency, followed by the SA and then CA. Interestingly, the performance of some particular CA strategies, such as the betweenness based attacks (BA), are even worse than the RDA. Furthermore, we discover that a link prediction method with high performance probably has lower attack robustness, and the vice versa.

physics.soc-ph↗

arXiv2018

Semidefinite Relaxation Based Blind Equalization using Constant Modulus Criterion

Blind equalization is a classic yet open problem. Statistic-based algorithms, such as constant modulus (CM), were widely investigated. One inherent issue with blind algorithms is the phase ambiguity of equalized signals. In this letter, we propose a novel scheme based on CM criterion and take advantage of the asymmetric property in a class of LDPC codes to resolve the phase ambiguity. Specifically, a new formulation with modified CM objective function and relaxed code constraints is presented.

cs.IT↗