arXiv subjects

Jin Li

Publications and source records attributed to Jin Li.

At least 271 records · Page 15Linked to original sources

arXiv2019

Cosmic opacity: cosmological-model-independent tests from gravitational waves and Type Ia Supernova

In this paper, we present a scheme to investigate the opacity of the Universe in a cosmological-model-independent way, with the combination of current and future available data in gravitational wave (GW) and electromagnetic (EM) domain. In the FLRW metric, GWs propagate freely through a perfect fluid without any absorption and dissipation, which provides a distance measurement unaffected by the cosmic opacity. Focusing on the simulated data of gravitational waves from the third-generation gravitational wave detector (the Einstein Telescope, ET), as well as the newly-compiled SNe Ia data (JLA and Pantheon sample), we find an almost transparent universe is strongly favored at much higher redshifts ($z\sim 2.26$). Our results suggest that, although the tests of cosmic opacity are not significantly sensitive to its parametrization, a strong degeneracy between the cosmic opacity parameter and the absolute \textit{B}-band magnitude of SNe Ia is revealed in this analysis. More importantly, we obtain that future measurements of the luminosity distances of gravitational waves sources will be much more competitive than the current analyses, which makes it expectable more vigorous and convincing constraints on the cosmic opacity (and consequently on background physical mechanisms) and a deeper understanding of the intrinsic properties of type Ia supernovae in a cosmological-model-independent way.

astro-ph.CO↗

arXiv2019

Channel adversarial training for cross-channel text-independent speaker recognition

The conventional speaker recognition frameworks (e.g., the i-vector and CNN-based approach) have been successfully applied to various tasks when the channel of the enrolment dataset is similar to that of the test dataset. However, in real-world applications, mismatch always exists between these two datasets, which may severely deteriorate the recognition performance. Previously, a few channel compensation algorithms have been proposed, such as Linear Discriminant Analysis (LDA) and Probabilistic LDA. However, these methods always require the collections of different channels from a specific speaker, which is unrealistic to be satisfied in real scenarios. Inspired by domain adaptation, we propose a novel deep-learning based speaker recognition framework to learn the channel-invariant and speaker-discriminative speech representations via channel adversarial training. Specifically, we first employ a gradient reversal layer to remove variations across different channels. Then, the compressed information is projected into the same subspace by adversarial training. Experiments on test datasets with 54,133 speakers demonstrate that the proposed method is not only effective at alleviating the channel mismatch problem, but also outperforms state-of-the-art speaker recognition methods. Compared with the i-vector-based method and the CNN-based method, our proposed method achieves significant relative improvement of 44.7% and 22.6% respectively in terms of the Top1 recall.

eess.AS↗

arXiv2019

Applying deep neural networks to the detection and space parameter estimation of compact binary coalescence with a network of gravitational wave detectors

In this paper, we study an application of deep learning to the advanced LIGO and advanced Virgo coincident detection of gravitational waves (GWs) from compact binary star mergers. This deep learning method is an extension of the Deep Filtering method used by George and Huerta (2017) for multi-inputs of network detectors. Simulated coincident time series data sets in advanced LIGO and advanced Virgo detectors are analyzed for estimating source luminosity distance and sky location. As a classifier, our deep neural network (DNN) can effectively recognize the presence of GW signals when the optimal signal-to-noise ratio (SNR) of network detectors $\geq$ 9. As a predictor, it can also effectively estimate the corresponding source space parameters, including the luminosity distance $D$, right ascension $α$, and declination $δ$ of the compact binary star mergers. When the SNR of the network detectors is greater than 8, their relative errors are all less than 23%. Our results demonstrate that Deep Filtering can process coincident GW time series inputs and perform effective classification and multiple space parameter estimation. Furthermore, we compare the results obtained from one, two, and three network detectors; these results reveal that a larger number of network detectors results in a better source location.

astro-ph.IM↗

arXiv2019

Testing the Etherington's distance duality relation at higher redshifts: the combination of radio quasars and gravitational waves

In this paper we analyse the implications of the latest cosmological data sets to test the Etherington's distance duality relation (DDR), which connects the luminosity distance $D_L$ and angular diameter distance $D_A$ at the same redshift. For $D_L$ we consider the simulated data of gravitational waves from the third-generation gravitational wave detector (the Einstein Telescope, ET), which can be considered as standard candles (or standard siren), while the angular diameter distances $D_A$ are derived from the newly-compiled sample of compact radio quasars observed by very-long-baseline interferometry (VLBI), which represents a type of new cosmological standard ruler. Alleviating the absorption and scattering effects of dust in the Universe, this will create a valuable opportunity to directly test DDR at much higher precision with the combination of gravitational wave (GW) and electromagnetic (EM) signals. Our results show that, with the combination of the current radio quasar observations, the duality-distance relation can be verified at the precision of $10^{-2}$. Moreover, the Einstein Telescope ET would produce more robust constraints on the validity of such distance duality relation (at the precision of $10^{-3}$), with a larger sample of compact milliarcsecond radio quasars detected in future VLBI surveys.

astro-ph.CO↗

arXiv2019

Flexible Ultrastrong 100-nm Polyethylene Membranes with Polygonal Pore Structures

Robust nanoporous polymer films with approximate 100-nm thicknesses would be broadly applicable in technological areas such as flexible sensors, artificial skins, separators, antireflection and self-cleaning films. However, the creation of these films has been extremely challenging. To date, all reported ultrathin films are of insufficient mechanical strength for use without substrate supports. We describe here the fabrication of a new ultrastrong and highly flexible ultrahigh molecular weight polyethylene (UHMWPE) 100-nm porous membrane. Scanning and transmission electron microscopy evaluations of film microstructures reveal a planar fibrous structure with randomly oriented polygonal interconnected pores. Moreover, the tensile strength of this film is twice as strong as that of solid stainless steel, with a measured in-plane tensile strength and ductility of approximately 900 MPa and 26%, respectively. We further coated our newly developed film with a monolayer graphene to invent an optically transparent and fully conformable piezoresistive skin sensor and thus demonstrated a potential application.

physics.app-ph↗

arXiv2019

Stone-Wales graphene: A Two Dimensional Carbon Semi-Metal with Magic Stability

A two-dimensional carbon allotrope, Stone-Wales graphene, is identified in stochastic group and graph constrained searches and systematically investigated by first-principles calculations. Stone-Wales graphene consists of well-arranged Stone-Wales defects, and it can be constructed through a 90$^\circ$ bond-rotation in a $\sqrt{8}$$\times$$\sqrt{8}$ super-cell of graphene. Its calculated energy relative to graphene, +149 meV/atom, makes it more stable than the most competitive previously suggested graphene allotropes. We find that Stone-Wales graphene based on a $\sqrt{8}$ super-cell is more stable than those based on $\sqrt{9} \times \sqrt{9}$, $\sqrt{12} \times \sqrt{12}$ and $\sqrt{13} \times \sqrt{13}$ super-cells, and is a "magic size" that can be further understood through a simple "energy splitting and inversion" model. The calculated vibrational properties and molecular dynamics of SW-graphene confirm that it is dynamically stable. The electronic structure shows SW-graphene is a semimetal with distorted, strongly anisotropic Dirac cones.

cond-mat.mtrl-sci↗

arXiv2018

Double Adaptive Stochastic Gradient Optimization

Adaptive moment methods have been remarkably successful in deep learning optimization, particularly in the presence of noisy and/or sparse gradients. We further the advantages of adaptive moment techniques by proposing a family of double adaptive stochastic gradient methods~\textsc{DASGrad}. They leverage the complementary ideas of the adaptive moment algorithms widely used by deep learning community, and recent advances in adaptive probabilistic algorithms.We analyze the theoretical convergence improvements of our approach in a stochastic convex optimization setting, and provide empirical validation of our findings with convex and non convex objectives. We observe that the benefits of~\textsc{DASGrad} increase with the model complexity and variability of the gradients, and we explore the resulting utility in extensions of distribution-matching multitask learning.

stat.ML↗

arXiv2018

Directional fast neutron detection using a time projection chamber and plastic scintillation detectors

A new method for directional fast neutron detection is proposed based on a neutron time projection chamber (TPC) and position-sensitive plastic scintillation detectors. The detection system can efficiently locate the approximate location of a hot spot with 4π field-of-view using only the neutron TPC. Then, the system generates a high-resolution image of the hot spot using selected coincidence events in the TPC and the scintillation detectors. A prototype was built and tested using a Cf-252 source. An efficiency of 7.1x10-3 was achieved for fast searching. The angular resolution was 7.8° (full width at half maximum, FWHM) for high-resolution imaging using the simple back projection method.

physics.ins-det↗

arXiv2018

Security Matters: A Survey on Adversarial Machine Learning

Adversarial machine learning is a fast growing research area, which considers the scenarios when machine learning systems may face potential adversarial attackers, who intentionally synthesize input data to make a well-trained model to make mistake. It always involves a defending side, usually a classifier, and an attacking side that aims to cause incorrect output. The earliest studies on the adversarial examples for machine learning algorithms start from the information security area, which considers a much wider varieties of attacking methods. But recent research focus that popularized by the deep learning community places strong emphasis on how the "imperceivable" perturbations on the normal inputs may cause dramatic mistakes by the deep learning with supposed super-human accuracy. This paper serves to give a comprehensive introduction to a range of aspects of the adversarial deep learning topic, including its foundations, typical attacking and defending strategies, and some extended studies.

cs.LG↗

arXiv2018

Complex low energy tetrahedral polymorphs of group IV elements from first-principles

The energy landscape of carbon is exceedingly complex, hosting diverse and important metastable phases, including diamond, fullerenes, nanotubes and graphene. Searching for structures, especially those with large unit cells, in this landscape is challenging. Here we use a combined stochastic search strategy employing two algorithms (AIRSS and RG2) to apply connectivity constraints to unit cells containing up to 100 carbon atoms. We uncover three low energy carbon polymorphs (Pbam-32, P6/mmm and I-43d) with new topologies, containing 32, 36 and 94 atoms in their primitive cells, respectively. Their energies relative to diamond are 96 meV/atom, 131 meV/atom and 112 meV/atom, respectively, which suggests potential metastability. These three carbon allotropes are mechanically and dynamically stable, insulating carbon crystals with superhard mechanical properties. The I43d structure possesses a direct band gap of 7.25 eV, which is the widest gap in the carbon allotrope family. Silicon, germanium and tin versions of Pbam-32, P6/mmm and I-43d also show energetic, dynamical and mechanical stability. The computed electronic properties show that they are potential materials for semiconductor and photovoltaic applications.

cond-mat.mtrl-sci↗

arXiv2018

A Real-time Robotic Grasp Approach with Oriented Anchor Box

Grasp is an essential skill for robots to interact with humans and the environment. In this paper, we build a vision-based, robust and real-time robotic grasp approach with fully convolutional neural network. The main component of our approach is a grasp detection network with oriented anchor boxes as detection priors. Because the orientation of detected grasps is significant, which determines the rotation angle configuration of the gripper, we propose the Orientation Anchor Box Mechanism to regress grasp angle based on predefined assumption instead of classification or regression without any priors. With oriented anchor boxes, the grasps can be predicted more accurately and efficiently. Besides, to accelerate the network training and further improve the performance of angle regression, Angle Matching is proposed during training instead of Jaccard Index Matching. Five-fold cross validation results demonstrate that our proposed algorithm achieves an accuracy of 98.8% and 97.8% in image-wise split and object-wise split respectively, and the speed of our detection algorithm is 67 FPS with GTX 1080Ti, outperforming all the current state-of-the-art grasp detection algorithms on Cornell Dataset both in speed and accuracy. Robotic experiments demonstrate the robustness and generalization ability in unseen objects and real-world environment, with the average success rate of 90.0% and 84.2% of familiar things and unseen things respectively on Baxter robot platform.

cs.RO↗

arXiv2018

An SOA Based Design of JUNO DAQ Online Software

The Online Software, manager of the JUNO data acquisition (DAQ) system, is composed of many distributed components working coordinately. It takes the responsibility of configuring, processes management, controlling and information sharing etc. The design of service-oriented architecture (SOA) which represents the modern tendency in the distributed system makes the online software lightweight, loosely coupled, reusable, modular, self-contained and easy to be extended. All the services in the SOA distributed online software system will send messages each to another directly without a traditional broker in the middle, which means that services could operate harmoniously and independently. ZeroMQ is chosen but not the only technical choice as the low-level communication middle-ware because of its high performance and convenient communication model while using Google Protocol Buffers as a marshaling library to unify the pattern of message contents. Considering the general requirement of JUNO, the concept of partition and segment are defined to ensure multiple small-scaled DAQs could run simultaneous and easy to join or leave. All running data except the raw physics events will be transmitted, processed and recorded to the database. High availability (HA) is also taken into account to solve the inevitable single point of failure (SPOF) in the distribution system. This paper will introduce all the core services' functionality and techniques in detail.

cs.SE↗

arXiv2018

Reversible Codes and Its Application to Reversible DNA Codes over $F_{4^k}$

Coterm polynomials are introduced by Oztas et al. [a novel approach for constructing reversible codes and applications to DNA codes over the ring $F_2[u]/(u^{2k}-1)$, Finite Fields and Their Applications 46 (2017).pp. 217-234.], which generate reversible codes. In this paper, we generalize the coterm polynomials and construct some reversible codes which are optimal codes by using $m$-quasi-reciprocal polynomials. Moreover, we give a map from DNA $k$-bases to the elements of $F_{4^k}$, and construct reversible DNA codes over $F_{4^k}$ by DNA-$m$-quasi-reciprocal polynomials.

cs.IT↗

arXiv2018

A Model on Heat Signal of Crystal Detector at Low Temperature

We present a model to calculate heat signal shapes from low temperature bolometer attached to a crystal. This model is based on the elementary acoustic wave theory at low temperature, and has been developed using modern Monte Carlo techniques. Physical processes in phonon propagation, such as transmission, scattering and reflection are considered. Using our model, the calculated time dependence of signal agrees with real experimental data. This model has applications in low temperature rare event particle detectors for dark matter and neutrinos.

physics.ins-det↗

arXiv2018

Testing the anisotropy of the universe using the simulated gravitational wave events from advanced LIGO and Virgo

The detection of gravitational waves (GWs) provides a powerful tool to constrain the cosmological parameters. In this paper, we investigate the possibility of using GWs as standard sirens in testing the anisotropy of the universe. We consider the GW signals produced by the coalescence of binary black hole systems and simulate hundreds of GW events from the advanced Laser Interferometer Gravitational-Wave Observatory (LIGO) and Virgo. It is found that the anisotropy of the universe can be tightly constrained if the redshift of the GW source is precisely known. The anisotropic amplitude can be constrained with an accuracy comparable to the Union2.1 complication of type-Ia supernovae if $\gtrsim 400$ GW events are observed. As for the preferred direction, $\gtrsim 800$ GW events are needed in order to achieve the accuracy of Union2.1. With 800 GW events, the probability of pseudo anisotropic signals with an amplitude comparable to Union2.1 is negligible. These results show that GWs can provide a complementary tool to supernovae in testing the anisotropy of the universe.

astro-ph.CO↗

arXiv2018

The Flash ADC system and PMT waveform reconstruction for the Daya Bay Experiment

To better understand the energy response of the Antineutrino Detector (AD), the Daya Bay Reactor Neutrino Experiment installed a full Flash ADC readout system on one AD that allowed for simultaneous data taking with the current readout system. This paper presents the design, data acquisition, and simulation of the Flash ADC system, and focuses on the PMT waveform reconstruction algorithms. For liquid scintillator calorimetry, the most critical requirement to waveform reconstruction is linearity. Several common reconstruction methods were tested but the linearity performance was not satisfactory. A new method based on the deconvolution technique was developed with 1% residual non-linearity, which fulfills the requirement. The performance was validated with both data and Monte Carlo (MC) simulations, and 1% consistency between them has been achieved.

physics.ins-det↗

arXiv2018

$L_p$ Minkowski Valuations on polytopes

For $1 \leq p < \infty$, Ludwig, Haberl and Parapatits classified $L_p$ Minkowski valuations intertwining the special linear group with additional conditions such as homogeneity and continuity. In this paper,a complete classification of $L_p$ Minkowski valuations intertwining the special linear group on polytopes without any additional conditions is established for $p \geq 1$ including $p = \infty$. For $n=3$ and $p=1$, there exist valuations not mentioned before.

math.MG↗

arXiv2018

Laplace transforms and valuations

It is proved that the classical Laplace transform is a continuous valuation which is positively GL$(n)$ covariant and logarithmic translation covariant. Conversely, these properties turn out to be sufficient to characterize this transform.

math.MG↗