arXiv subjects

Bo Qin

Publications and source records attributed to Bo Qin.

At least 37 records · Page 2Linked to original sources

arXiv2022

Handwritten Mathematical Expression Recognition via Attention Aggregation based Bi-directional Mutual Learning

Handwritten mathematical expression recognition aims to automatically generate LaTeX sequences from given images. Currently, attention-based encoder-decoder models are widely used in this task. They typically generate target sequences in a left-to-right (L2R) manner, leaving the right-to-left (R2L) contexts unexploited. In this paper, we propose an Attention aggregation based Bi-directional Mutual learning Network (ABM) which consists of one shared encoder and two parallel inverse decoders (L2R and R2L). The two decoders are enhanced via mutual distillation, which involves one-to-one knowledge transfer at each training step, making full use of the complementary information from two inverse directions. Moreover, in order to deal with mathematical symbols in diverse scales, an Attention Aggregation Module (AAM) is proposed to effectively integrate multi-scale coverage attentions. Notably, in the inference phase, given that the model already learns knowledge from two inverse directions, we only use the L2R branch for inference, keeping the original parameter size and inference speed. Extensive experiments demonstrate that our proposed approach achieves the recognition accuracy of 56.85 % on CROHME 2014, 52.92 % on CROHME 2016, and 53.96 % on CROHME 2019 without data augmentation and model ensembling, substantially outperforming the state-of-the-art methods. The source code is available in https://github.com/XH-B/ABM.

cs.CV↗

arXiv2020

The conformity of HI galaxies in ALFALFA-SDSS sample

The conformity effect, indicating the evolution of a galaxy is related to its surrounding neighbour galaxies as far as a few Mpc, is an interesting phenomenon in the modeling of galaxy and evolution. Here we study the conformity effect of HI galaxies in a matched galaxy sample between SDSS DR7 and ALFALFA surveys. By checking the probability difference for the detected HI galaxies as a function of distance around a normal or an HI galaxy, we find that this effect is significant out to 5 Mpc. It also shows a dependence on the stellar mass of galaxies, with the strength the strongest in the stellar mass range of 10^10~10^10.5. However, when the sample is confined to central galaxies in groups with virial radii smaller than 1 Mpc, the 1-halo conformity signal is still evident, while the 2-halo conformity signal is reduced to a very weak amplitude. Our results confirm the previous study in the optical band that the 2-halo term is possibly caused by the bias effect in the sample selection. Furthermore, we confirm the existence of the 1-halo conformity discovered in the optical and radio band in previous lectures. Our results provide another viewpoint of the conformity effect and hope to shed the light on the co-evolution of the galaxies and their neighbours in the current galaxy formation models.

astro-ph.GA↗

arXiv2019

The Giant Radio Array for Neutrino Detection (GRAND): Science and Design

The Giant Radio Array for Neutrino Detection (GRAND) is a planned large-scale observatory of ultra-high-energy (UHE) cosmic particles, with energies exceeding 10^8 GeV. Its goal is to solve the long-standing mystery of the origin of UHE cosmic rays. To do this, GRAND will detect an unprecedented number of UHE cosmic rays and search for the undiscovered UHE neutrinos and gamma rays associated to them with unmatched sensitivity. GRAND will use large arrays of antennas to detect the radio emission coming from extensive air showers initiated by UHE particles in the atmosphere. Its design is modular: 20 separate, independent sub-arrays, each of 10 000 radio antennas deployed over 10 000 km^2. A staged construction plan will validate key detection techniques while achieving important science goals early. Here we present the science goals, detection strategy, preliminary design, performance goals, and construction plans for GRAND.

astro-ph.HE↗

arXiv2017

Non-approximability and Polylogarithmic Approximations of the Single-Sink Unsplittable and Confluent Dynamic Flow Problems

Dynamic Flows were introduced by Ford and Fulkerson in 1958 to model flows over time. They define edge capacities to be the total amount of flow that can enter an edge {\em in one time unit}. Each edge also has a length, representing the time needed to traverse it. Dynamic Flows have been used to model many problems including traffic congestion, hop-routing of packets and evacuation protocols in buildings. While the basic problem of moving the maximal amount of supplies from sources to sinks is polynomial time solvable, natural minor modifications can make it NP-hard. One such modification is that flows be confluent, i.e., all flows leaving a vertex must leave along the same edge. This corresponds to natural conditions in, e.g., evacuation planning and hop routing. We investigate the single-sink Confluent Quickest Flow problem. The input is a graph with edge capacities and lengths, sources with supplies and a sink. The problem is to find a confluent flow minimizing the time required to send supplies to the sink. Our main results include: a) Logarithmic Non-Approximability: Directed Confluent Quickest Flows cannot be approximated in polynomial time with an $O(\log n)$ approximation factor, unless $P=NP$. b) Polylogarithmic Bicriteria Approximations: Polynomial time $(O(\log^8 n), O(\log^2 κ))$ bicritera approximation algorithms for the Confluent Quickest Flow problem where $κ$ is the number of sinks, in both directed and undirected graphs. Corresponding results are also developed for the Confluent Maximum Flow over timeproblem. The techniques developed also improve recent approximation algorithms for static confluent flows.

cs.DS↗

arXiv2015

Flexible Attribute-Based Encryption Applicable to Secure E-Healthcare Records

In e-healthcare record systems (EHRS), attribute-based encryption (ABE) appears as a natural way to achieve fine-grained access control on health records. Some proposals exploit key-policy ABE (KP-ABE) to protect privacy in such a way that all users are associated with specific access policies and only the ciphertexts matching the users' access policies can be decrypted. An issue with KP-ABE is that it requires an a priori formulation of access policies during key generation, which is not always practicable in EHRS because the policies to access health records are sometimes determined after key generation. In this paper, we revisit KPABE and propose a dynamic ABE paradigm, referred to as access policy redefinable ABE (APR-ABE). To address the above issue, APR-ABE allows users to redefine their access policies and delegate keys for the redefined ones; hence a priori precise policies are no longer mandatory. We construct an APR-ABE scheme with short ciphertexts and prove its full security in the standard model under several static assumptions.

cs.CR↗

arXiv2015

On Sketching Quadratic Forms

We undertake a systematic study of sketching a quadratic form: given an $n \times n$ matrix $A$, create a succinct sketch $\textbf{sk}(A)$ which can produce (without further access to $A$) a multiplicative $(1+ε)$-approximation to $x^T A x$ for any desired query $x \in \mathbb{R}^n$. While a general matrix does not admit non-trivial sketches, positive semi-definite (PSD) matrices admit sketches of size $Θ(ε^{-2} n)$, via the Johnson-Lindenstrauss lemma, achieving the "for each" guarantee, namely, for each query $x$, with a constant probability the sketch succeeds. (For the stronger "for all" guarantee, where the sketch succeeds for all $x$'s simultaneously, again there are no non-trivial sketches.) We design significantly better sketches for the important subclass of graph Laplacian matrices, which we also extend to symmetric diagonally dominant matrices. A sequence of work culminating in that of Batson, Spielman, and Srivastava (SIAM Review, 2014), shows that by choosing and reweighting $O(ε^{-2} n)$ edges in a graph, one achieves the "for all" guarantee. Our main results advance this front. $\bullet$ For the "for all" guarantee, we prove that Batson et al.'s bound is optimal even when we restrict to "cut queries" $x\in \{0,1\}^n$. In contrast, previous lower bounds showed the bound only for {\em spectral-sparsifiers}. $\bullet$ For the "for each" guarantee, we design a sketch of size $\tilde O(ε^{-1} n)$ bits for "cut queries" $x\in \{0,1\}^n$. We prove a nearly-matching lower bound of $Ω(ε^{-1} n)$ bits. For general queries $x \in \mathbb{R}^n$, we construct sketches of size $\tilde{O}(ε^{-1.6} n)$ bits.

cs.DS↗

arXiv2015

On the Security of Privacy-Preserving Vehicular Communication Authentication with Hierarchical Aggregation and Fast Response

In [3], the authors proposed a highly efficient secure and privacy-preserving scheme for secure vehicular communications. The proposed scheme consists of four protocols: system setup, protocol for STP and STK distribution, protocol for common string synchronization, and protocol for vehicular communications. Here we define the security models for the protocol for STP and STK distribution, and the protocol for vehicular communications,respectively. We then prove that these two protocols are secure in our models.

cs.CR↗

arXiv2015

On the Security of MTA-OTIBASs (Multiple-TA One-Time Identity-Based Aggregate Signatures)

In [3] the authors proposed a new aggregate signature scheme referred to as multiple-TA (trusted authority) one-time identity-based aggregate signature (MTA-OTIBAS). Further, they gave a concrete MTA-OTIBAS scheme. We recall here the definition of MTA-OTIBAS and the concrete proposed scheme. Then we prove that our MTA-OTIBAS concrete scheme is existentially unforgeable against adaptively chosen-message attacks in the random oracle model under the co-CDH problem assumption.

cs.CR↗

arXiv2014

A Sketching Algorithm for Spectral Graph Sparsification

We study the problem of compressing a weighted graph $G$ on $n$ vertices, building a "sketch" $H$ of $G$, so that given any vector $x \in \mathbb{R}^n$, the value $x^T L_G x$ can be approximated up to a multiplicative $1+ε$ factor from only $H$ and $x$, where $L_G$ denotes the Laplacian of $G$. One solution to this problem is to build a spectral sparsifier $H$ of $G$, which, using the result of Batson, Spielman, and Srivastava, consists of $O(n ε^{-2})$ reweighted edges of $G$ and has the property that simultaneously for all $x \in \mathbb{R}^n$, $x^T L_H x = (1 \pm ε) x^T L_G x$. The $O(n ε^{-2})$ bound is optimal for spectral sparsifiers. We show that if one is interested in only preserving the value of $x^T L_G x$ for a {\it fixed} $x \in \mathbb{R}^n$ (specified at query time) with high probability, then there is a sketch $H$ using only $\tilde{O}(n ε^{-1.6})$ bits of space. This is the first data structure achieving a sub-quadratic dependence on $ε$. Our work builds upon recent work of Andoni, Krauthgamer, and Woodruff who showed that $\tilde{O}(n ε^{-1})$ bits of space is possible for preserving a fixed {\it cut query} (i.e., $x\in \{0,1\}^n$) with high probability; here we show that even for a general query vector $x \in \mathbb{R}^n$, a sub-quadratic dependence on $ε$ is possible. Our result for Laplacians is in sharp contrast to sketches for general $n \times n$ positive semidefinite matrices $A$ with $O(\log n)$ bit entries, for which even to preserve the value of $x^T A x$ for a fixed $x \in \mathbb{R}^n$ (specified at query time) up to a $1+ε$ factor with constant probability, we show an $Ω(n ε^{-2})$ lower bound.

cs.DS↗

arXiv2010

Mass discrepancy in galaxy clusters as a result of the offset between dark matter and baryon distributions

Recent studies of lensing clusters reveal that it might be fairly common for a galaxy cluster that the X-ray center has an obvious offset from its gravitational center which is measured by strong lensing. We argue that if these offsets exist, then X-rays and lensing are indeed measuring different regions of a cluster, and may thus naturally result in a discrepancy in the measured gravitational masses by the two different methods. Here we investigate theoretically the dynamical effects of such lensing-X-ray offsets, and compare with observational data. We find that for typical values, the offset alone can give rise to a factor of two difference between the lensing and X-ray determined masses for the core regions of a cluster, suggesting that such "offset effect" may play an important role and should not be ignored in our dynamical measurements of clusters.

astro-ph.CO↗

arXiv2010

Offset between dark matter and ordinary matter: evidence from a sample of 38 lensing clusters of galaxies

We compile a sample of 38 galaxy clusters which have both X-ray and strong lensing observations, and study for each cluster the projected offset between the dominant component of baryonic matter center (measured by X-rays) and the gravitational center (measured by strong lensing). Among the total sample, 45% clusters have offsets >10". The >10" separations are significant, considering the arcsecond precision in the measurement of the lensing/X-ray centers. This suggests that it might be a common phenomenon in unrelaxed galaxy clusters that gravitational field is separated spatially from the dominant component of baryonic matter. It also has consequences for lensing models of unrelaxed clusters since the gas mass distribution may differ from the dark matter distribution and give perturbations to the modeling. Such offsets can be used as a statistical tool for comparison with the results of Lambda-CDM simulations and to test the modified dynamics.

astro-ph.CO↗

arXiv2008

Galaxy Distribution as a Probe of the Ringlike Dark Matter Structure in the Galaxy Cluster Cl0024+17

In a galaxy cluster, galaxies are mostly collisionless particles in recent epoches. They resemble collisionless cold dark matter particles in some way. Therefore, the spatial distributions of dark matter and cluster galaxies might be expected to possess similar features in the gravitational potential of a cluster. Here we use the galaxy distribution in cluster Cl0024+17 to probe for the ringlike dark matter structure recently discovered by means of strong and weak lensing observations. The galaxies are taken from the catalog of Czoske et al., which contains 650 objects with measured redshifts, of which ~300 galaxies have redshifts in the range 0.37<z<0.41 (and are therefore probable cluster members). We find that, at about the 3-sigma level, the ringlike structure seen in the dark matter measurement is not observed in the projected two-dimensional galaxy distribution.

astro-ph↗

arXiv2007

The Effects of Metallicity and AGN Activity on the Mid-Infrared Dust Emission of Galaxies

Using a sample of the $Spitzer$ SWIRE-field galaxies whose optical spectra are taken from the Data Release 4 of the Sloan Digital Sky Survey, we study possible correlations between the Mid-InfraRed (MIR) dust emission from these galaxies and both their metallicities and AGN activities. We find that both metallicity and AGN activity are well correlated with the following ratios: PAH(Polycyclic Aromatic Hydrocarbon)-to-star, VSG(Very Small Grain)-to-star and PAH-to-VSG, which can be characterized by $νL_ν[8μm({\rm dust})]/νL_ν[3.6μm]$, $νL_ν[24μm]/νL_ν[3.6μm]$, and$νL_ν[8μm({\rm dust})]/νL_ν[24μm]$ respectively. We argue that our MIR-metallicity correlation could be explained by either the amount of dust (ongoing dust formation) or dust destruction (PAHs and VSGs could be destroyed by hard and intense radiation fields), and that the MIR-AGN correlation could arise due to either PAH destruction or an enhanced VSG continuum by the central AGN.

astro-ph↗

arXiv2005

Observational Evidence for Extra Dimensions from Dark Matter

Recent astronomical observations of systems of dark matter, which have been cited as providing possible support for self-interacting cold dark matter, may provide evidence for the extra dimensions predicted by superstring scenarios. We find that the properties of the required dark matter self-interaction are precisely the consequences of a world with 3 large extra dimensions of size \~1nm, where gravity follows the r^{-5} law at scales below ~1nm. From the cross sections measured for various dark matter systems, we also constrain the mass of dark matter particles to be m_x ~ 3*10^{-16} proton mass, consistent with the mass of axions.

astro-ph↗

arXiv2002

A New Mass Degeneracy in Gravitational Lenses: Is there a crisis between a large $H_0$ and the time delay by a large lens halo?

Isothermal models and other simple smooth models of dark matter halos of gravitational lenses often predict a dimensionless time delay $H_0Δt$ much too small to be comfortable with the observed time delays $Δt$ and the widely accepted $H_0$ value ($\sim 70$ km/s/Mpc). This conflict or crisis of the CDM has been highlighted by several recent papers of Kochanek, who claims that the standard value of $H_0$ favors a strangely small halo as compact as the stellar light distribution with an overall nearly Keplerian rotation curve. In an earlier paper (Paper I, astro-ph/0209191) we argue that this is not necessarily the case, at least in a perfectly symmetrical Einstein cross system. Here we introduce a {\it new mass degeneracy} of lens systems to give a counter example to Kochanek's claims. We fit the time delay and image positions in the quadruple image system PG1115+080. Equally good fits are found between lens models with flat vs. Keplerian rotation curves. Time delays in both types of models can be fit with the standard value of $H_0$. We demonstrate that it may still be problematic to constrain the size of lens dark halos even if the data image positions are accurately given and the cosmology is precisely specified.

astro-ph↗

arXiv2002

Intrinsic Degeneracy of Gravitational Lens Time Delays: the case for a simple quadruple system in $Λ$CDM cosmology

A degeneracy in strong lens model is shown analytically. The observed time delays and quasar image positions might {\it not} uniquely determine the concentration and the extent of the lens galaxy halo mass distribution. Simply hardwiring the Hubble constant ($H_0$) and the cosmology ($Ω, Λ$) to the standard values of $Λ$CDM cosmology might {\it not} fully lift this degeneracy, which exists rigourously even with very accurate data. Equally good fits to the images could be found in lens mass models with either a mostly Keplerian or a flat rotation curve. This degeneracy in mass models makes the task of getting reliable $H_0$ and $Λ$ from strong lenses even more daunting.

astro-ph↗

arXiv2001

Constraints on the Interaction between Dark Matter and Baryons from Cooling Flow Clusters

Other nongravitational heating processes are needed to resolve the disagreement between the absence of cool gas components in the centers of galaxy clusters revealed recently by Chandra and XMM observations and the expectations of conventional radiative cooling models. Here we propose that the interaction between dark matter particles and ordinary baryonic matter may act as an alternative for the reheating of intracluster medium (ICM) in the inner regions of clusters, in which kinetic energy of dark matter is transported to ICM to balance the radiative cooling. Using the Chandra and XMM data of typical clusters, we set a useful constraint on the dark matter-baryon cross-section: $σ_{xp}/m_x \sim 1 \times 10^{-25}$ cm$^2$GeV$^{-1}$, where $m_x$ is the mass of dark matter particles.

astro-ph↗

arXiv1999

Baryon Distribution in Galaxy Clusters as a Result of Sedimentation of Helium Nuclei

Heavy particles in galaxy clusters tend to be more centrally concentrated than light ones according to the Boltzmann distribution. An estimate of the drift velocity suggests that it is possible that the helium nuclei may have entirely or partially sedimented into the cluster core within the Hubble time. We demonstrate the scenario using the NFW profile as the dark matter distribution of clusters and assuming that the intracluster gas is isothermal and in hydrostatic equilibrium. We find that a greater fraction of baryonic matter is distributed at small radii than at large radii, which challenges the prevailing claim that the baryon fraction increases monotonically with cluster radius. It shows that the conventional mass estimate using X-ray measurements of intracluster gas along with a constant mean molecular weight may have underestimated the total cluster mass by $\sim 20%$, which in turn leads to an overestimate of the total baryon fraction by the same percentage. Additionally, it is pointed out that the sedimentation of helium nuclei toward cluster cores may at least partially account for the sharp peaks in the central X-ray emissions observed in some clusters.

astro-ph↗