arXiv subjects

Bin Liu

Publications and source records attributed to Bin Liu.

At least 469 records · Page 26Linked to original sources

arXiv2018

Enhanced Black-Hole Mergers in Binary-Binary Interactions

We study the orbital evolution of black hole (BH) binaries in quadruple systems, where the tertiary binary excites large eccentricity in the BH binary through Lidov-Kozai (LK) oscillations, causing the binary BHs to merge via gravitational radiation. For typical BH binaries with masses $m_{1,2}\simeq 20M_\odot-30M_\odot$ and initial semimajor axis $a_0\sim100$ AU (such that the binaries have no chance of merging by themselves within $\sim10^{10}$ yrs), we show that binary-binary interactions can significantly increase the LK window for mergers (the range of companion inclinations that allows the BH binary to merge within 10~Gyrs). This increase arises from a secular resonance between the LK oscillation of the BH binary and the nodal precession of the outer (binary-binary) orbit driven by the tertiary binary. Therefore, in the presence of tertiary binary, the BH merger fraction is increased to $10-30\%$, an order of magnitude larger than the merger fraction found in similar triple systems. Our analysis (with appropriate scalings) can be easily adapted to other configurations of systems, such as relatively compact BH binaries and moderately hierarchical triples, which may generate even higher merger fractions. Since the occurrence rate of stellar quadruples in the galactic fields is not much smaller than that of stellar triples, our result suggests that dynamically induced BH mergers in quadruple systems may be an important channel of producing BH mergers observed by LIGO/VIRGO.

astro-ph.HE↗

arXiv2018

Real-Time Anomaly Detection With HMOF Feature

Anomaly detection is a challenging problem in intelligent video surveillance. Most existing methods are computation consuming, which cannot satisfy the real-time requirement. In this paper, we propose a real-time anomaly detection framework with low computational complexity and high efficiency. A new feature, named Histogram of Magnitude Optical Flow (HMOF), is proposed to capture the motion of video patches. Compared with existing feature descriptors, HMOF is more sensitive to motion magnitude and more efficient to distinguish anomaly information. The HMOF features are computed for foreground patches, and are reconstructed by the auto-encoder for better clustering. Then, we use Gaussian Mixture Model (GMM) Classifiers to distinguish anomalies from normal activities in videos. Experimental results show that our framework outperforms state-of-the-art methods, and can reliably detect anomalies in real-time.

cs.CV↗

arXiv2018

Ionization Profiles of Galactic HII Regions

Using Green Bank Telescope radio recombination line (RRL) data, we analyze the role of leaking radiation from HII regions in maintaining the ionization of the interstellar medium. We observed a sample of eight Galactic HII regions of various sizes, morphologies, and luminosities. For each region the hydrogen RRL intensity decreases roughly as a power-law with distance from the center of the region. This suggests that radiation leaking from the HII region is responsible for the majority of surrounding ionized gas producing RRL emission. Our results further indicate that the hydrogen RRL intensity appears to be fundamentally related to the HII region sizes traced by their photodissociation regions, such that physically smaller HII regions show a steeper decrease in intensity with increasing distance from the region centers. As a result, giant HII regions may have a much larger effect in maintaining the ionization of the interstellar medium. For six of the eight observed HII regions we find a decrease in the 4He+/H+ abundance ratio with increasing distance, indicating that He-ionizing photons are being absorbed within the ionization front of the HII region. There is enhanced carbon RRL emission toward directions with strong continuum background, suggesting that the carbon emission is amplified by stimulated emission.

astro-ph.GA↗

arXiv2018

What and Where: A Context-based Recommendation System for Object Insertion

In this work, we propose a novel topic consisting of two dual tasks: 1) given a scene, recommend objects to insert, 2) given an object category, retrieve suitable background scenes. A bounding box for the inserted object is predicted in both tasks, which helps downstream applications such as semi-automated advertising and video composition. The major challenge lies in the fact that the target object is neither present nor localized at test time, whereas available datasets only provide scenes with existing objects. To tackle this problem, we build an unsupervised algorithm based on object-level contexts, which explicitly models the joint probability distribution of object categories and bounding boxes with a Gaussian mixture model. Experiments on our newly annotated test set demonstrate that our system outperforms existing baselines on all subtasks, and do so under a unified framework. Our contribution promises future extensions and applications.

cs.CV↗

arXiv2018

A Particle Filter based Multi-Objective Optimization Algorithm: PFOPS

This paper is concerned with a recently developed paradigm for population-based optimization, termed particle filter optimization (PFO). This paradigm is attractive in terms of coherence in theory and easiness in mathematical analysis and interpretation. Current PFO algorithms only work for single-objective optimization cases, while many real-life problems involve multiple objectives to be optimized simultaneously. To this end, we make an effort to extend the scope of application of the PFO paradigm to multi-objective optimization (MOO) cases. An idea called path sampling is adopted within the PFO scheme to balance the different objectives to be optimized. The resulting algorithm is thus termed PFO with Path Sampling (PFOPS). The validity of the presented algorithm is assessed based on three benchmark MOO experiments, in which the shapes of the Pareto fronts are convex, concave and discontinuous, respectively.

stat.ML↗

arXiv2018

A projection-based numerical integration scheme for embedded interface: Application to fluid-structure interaction

We present a projection-based numerical integration technique to deal with embedded interface in finite element (FE) framework. The element cut by an embedded interface is denoted as a cut cell. We recognize elemental matrices of a cut cell can be reconstructed from the elemental matrices of its sub-divided cells, via projection at matrix level. These sub-divided cells are termed as integration cells. The proposed technique possesses following characteristics (1) no change in FE formulation and quadrature rule; (2) consistency with the derivation of FE formulation in variational principle. It can be considered as a re-projection of the residuals of equation system in the test function space or a reduced-order modeling (ROM) technique. These characteristics significantly improves its scalability, easy-to-implementation and robustness to deal with problems involving embedded discontinuities in FE framework. Numerical examples, e.g., vortex-induced vibration (VIV), rotation, free fall and rigid-body contact in which the proposed technique is implemented to integrate the variational form of Navier-Stokes equations in cut cells, are presented.

math.NA↗

arXiv2018

Supercondutivity in SnSb with natural superlattice structure

We report the results of electrical resistivity, magnetic and thermodynamic measurements on polycrystalline SnSb, whose structure consists of stacks of Sb bilayers and Sn4Sb3 septuple layers along the c-axis. The material is found to be a weakly coupled, fully gapped, type-II superconductor with a bulk Tc of 1.50 K, while showing a zero resistivity transition at a significantly higher temperature of 2.48 K. The Sommerfeld coefficient and upper critical field, obtained from specific heat measurements, are 2.29 mJ/mol K and 520 Oe, respectively. Compositional inhomogeneity and strain effect at the grain boundaries are proposed as possible origins for the difference in resistive and bulk superconducting transitions.In addition, a comparison with the rock-salt structure SnAs superconductor is presented. Our results provide the first clear evidence of bulk superconductivity in a natural superlattice derived from a topological semimetal.

cond-mat.supr-con↗

arXiv2018

A Very Brief and Critical Discussion on AutoML

This contribution presents a very brief and critical discussion on automated machine learning (AutoML), which is categorized here into two classes, referred to as narrow AutoML and generalized AutoML, respectively. The conclusions yielded from this discussion can be summarized as follows: (1) most existent research on AutoML belongs to the class of narrow AutoML; (2) advances in narrow AutoML are mainly motivated by commercial needs, while any possible benefit obtained is definitely at a cost of increase in computing burdens; (3)the concept of generalized AutoML has a strong tie in spirit with artificial general intelligence (AGI), also called "strong AI", for which obstacles abound for obtaining pivotal progresses.

cs.AI↗

arXiv2018

Making Classifier Chains Resilient to Class Imbalance

Class imbalance is an intrinsic characteristic of multi-label data. Most of the labels in multi-label data sets are associated with a small number of training examples, much smaller compared to the size of the data set. Class imbalance poses a key challenge that plagues most multi-label learning methods. Ensemble of Classifier Chains (ECC), one of the most prominent multi-label learning methods, is no exception to this rule, as each of the binary models it builds is trained from all positive and negative examples of a label. To make ECC resilient to class imbalance, we first couple it with random undersampling. We then present two extensions of this basic approach, where we build a varying number of binary models per label and construct chains of different sizes, in order to improve the exploitation of majority examples with approximately the same computational budget. Experimental results on 16 multi-label datasets demonstrate the effectiveness of the proposed approaches in a variety of evaluation metrics.

cs.LG↗

arXiv2018

Deep Segment Attentive Embedding for Duration Robust Speaker Verification

LSTM-based speaker verification usually uses a fixed-length local segment randomly truncated from an utterance to learn the utterance-level speaker embedding, while using the average embedding of all segments of a test utterance to verify the speaker, which results in a critical mismatch between testing and training. This mismatch degrades the performance of speaker verification, especially when the durations of training and testing utterances are very different. To alleviate this issue, we propose the deep segment attentive embedding method to learn the unified speaker embeddings for utterances of variable duration. Each utterance is segmented by a sliding window and LSTM is used to extract the embedding of each segment. Instead of only using one local segment, we use the whole utterance to learn the utterance-level embedding by applying an attentive pooling to the embeddings of all segments. Moreover, the similarity loss of segment-level embeddings is introduced to guide the segment attention to focus on the segments with more speaker discriminations, and jointly optimized with the similarity loss of utterance-level embeddings. Systematic experiments on Tongdun and VoxCeleb show that the proposed method significantly improves robustness of duration variant and achieves the relative Equal Error Rate reduction of 50% and 11.54% , respectively.

eess.AS↗

arXiv2018

Investigating Continuous Power Flow Solutions of IEEE-14 Bus System

This letter focuses on the multiplicity of power flow (PF) equations and presents two continuous solutions for widely studied IEEE-14 bus system. The continuous solutions are located by a method combining the semidefinite program (SDP) relaxation and reformulation linearization technique (RLT). Although the observation is non-trivial, it is of interest to researchers investigating the geometry or multiplicity nature of PF equations.

math.OC↗

arXiv2018

Estimating Lower Probability Bound of Power System's Capability to Fully Accommodate Variable Wind Generation

As the penetration of wind generation increases, the uncertainty it brings has imposed great challenges to power system operation. To cope with the challenges, tremendous research work has been conducted, among which two aspects are of most importance, i.e. making immune operation strategies and accessing the power system's capability to accommodate the variable energy. Driven and inspired by the latter problem, this paper will discuss the power system's capability to accommodate variable wind generation in a probability sense. Wind generation, along with its uncertainty is illustrated by a polyhedron, which contains prediction, risk and uncertainty information. Then, a three-level optimization problem is presented to estimate the lower probability bound of power system's capability to fully accommodate wind generation. After reformulating the inner \emph{max-min} problem, or feasibility check problem, into its equivalent mixed-integer linear program (MILP) form, the bisection algorithm is presented to solve this challenging problem. Modified IEEE systems are adopted to show the effectiveness of the proposed method.

math.OC↗

arXiv2018

Dispersion of third-harmonic generation in ultrastrongly coupled organic cavity polaritons

Organic cavity polaritons are bosonic quasi-particles that arise from the strong interaction between organic molecular excitons and photons within microcavities. The spectral dispersion of third harmonic generation near resonance with the cavity polariton states is studied experimentally via angle-resolved reflected third harmonic generation measurements with several pump wavelengths. In addition, a three-step nonlinear optical transfer matrix model is used to simulate the third harmonic generation using the sum-over-states dispersive nonlinear coefficients, which include hybrid exciton-photon polariton states. The angle-dependent experiment and modeling agree, revealing that the output of third harmonic generation is resonantly enhanced when the third harmonic wavelength is near the upper polariton state. The degree of enhancement is higher on the exciton-like branch of the polariton dispersion. Lower polariton enhanced third harmonic generation is not experimentally observed due to inadequate coupling at the longer wavelength, which is also consistent with the nonlinear transfer matrix modeling. These results indicate that the sum-over-states nonlinear dispersion is descriptive of the process implying that the polariton states are terms in a complete set of states forming the basis for the perturbative calculation of the nonlinear optical response. The results also indicate the possibility of wavelength agile nonlinear optical response by angle-of-incidence tuning.

physics.optics↗

arXiv2018

Black Hole and Neutron Star Binary Mergers in Triple Systems: Merger Fraction and Spin-Orbit Misalignment

Black hole (BH) mergers driven by gravitational perturbations of external companions constitute an important class of formation channels for merging BH binaries detected by LIGO. We have studied the orbital and spin evolution of binary BHs in triple systems, where the tertiary companion excites large eccentricity in the inner binary through Lidov-Kozai oscillations, causing the binary to merge via gravitational radiation. Using the single-orbital-averaged and double-orbital-averaged secular dynamics equations of triples, we determine the merger window (the range of companion inclinations that allows the inner binary to merge) and the merger fraction as a function of various system parameters. For typical BH binaries, the merger fraction increases rapidly with the perturber eccentricity $e_{\rm out}$ because of the octupole perturbation, reaching $10-20\%$ at $e_{\rm out}=0.9$. We derive the analytical expressions for the merger window and merger fraction for systems with negligible octupole effect, and apply them to neutron star binary mergers in triples. We also follow the spin evolution of the BHs during the companion-induced orbital decay. Starting from aligned spin axes, a wide range of final spin-orbit misalignment angle $θ_{\rm sl}^{\rm f}$ can be generated when the binary enters the LIGO band. For systems where the octupole effect is small (such as equal mass binaries or $e_{\rm out}\sim 0$), the distribution of $θ_{\rm sl}^{\rm f}$ peaks around $90^\circ$. As the octuple effect increases, a more isotropic distribution of final spin axis is produced. Overall, merging BH binaries produced by Lidov-Kozai oscillations in triples exhibit a unique distribution of the effective spin parameter $χ_{\rm eff}$; this may be used to distinguish this formation channel from others.

astro-ph.HE↗

arXiv2018

Zoom-Net: Mining Deep Feature Interactions for Visual Relationship Recognition

Recognizing visual relationships among any pair of localized objects is pivotal for image understanding. Previous studies have shown remarkable progress in exploiting linguistic priors or external textual information to improve the performance. In this work, we investigate an orthogonal perspective based on feature interactions. We show that by encouraging deep message propagation and interactions between local object features and global predicate features, one can achieve compelling performance in recognizing complex relationships without using any linguistic priors. To this end, we present two new pooling cells to encourage feature interactions: (i) Contrastive ROI Pooling Cell, which has a unique deROI pooling that inversely pools local object features to the corresponding area of global predicate features. (ii) Pyramid ROI Pooling Cell, which broadcasts global predicate features to reinforce local object features.The two cells constitute a Spatiality-Context-Appearance Module (SCA-M), which can be further stacked consecutively to form our final Zoom-Net.We further shed light on how one could resolve ambiguous and noisy object and predicate annotations by Intra-Hierarchical trees (IH-tree). Extensive experiments conducted on Visual Genome dataset demonstrate the effectiveness of our feature-oriented approach compared to state-of-the-art methods (Acc@1 11.42% from 8.16%) that depend on explicit modeling of linguistic interactions. We further show that SCA-M can be incorporated seamlessly into existing approaches to improve the performance by a large margin. The source code will be released on https://github.com/gjyin91/ZoomNet.

cs.CV↗

arXiv2018

Stability band structure for periodic states in periodic potentials

A class of periodic solutions of the nonlinear Schrodinger equation with non- Hermitian potentials are considered. The system may be implemented in planar nonlinear optical waveguides carrying an appropriate distribution of local gain and loss, in a combination with a photonic-crystal structure. The complex potential is built as a solution of the inverse problem, which predicts the potential supporting required periodic solutions. The main subject of the analysis is the spectral structure of the linear (in)stability for the stationary spatially periodic states in the periodic potentials. The stability and instability bands are calculated by means of the plane-wave-expansion method, and verified in direct simulations of the perturbed evolution. The results show that the periodic solutions may be stable against perturbations in specific Floquet-Bloch bands, even if they are unstable against small random perturbations.

physics.optics↗

arXiv2018

Bayesian Methods for Analysis and Adaptive Scheduling of Exoplanet Observations

We describe work in progress by a collaboration of astronomers and statisticians developing a suite of Bayesian data analysis tools for extrasolar planet (exoplanet) detection, planetary orbit estimation, and adaptive scheduling of observations. Our work addresses analysis of stellar reflex motion data, where a planet is detected by observing the "wobble" of its host star as it responds to the gravitational tug of the orbiting planet. Newtonian mechanics specifies an analytical model for the resulting time series, but it is strongly nonlinear, yielding complex, multimodal likelihood functions; it is even more complex when multiple planets are present. The parameter spaces range in size from few-dimensional to dozens of dimensions, depending on the number of planets in the system, and the type of motion measured (line-of-sight velocity, or position on the sky). Since orbits are periodic, Bayesian generalizations of periodogram methods facilitate the analysis. This relies on the model being linearly separable, enabling partial analytical marginalization, reducing the dimension of the parameter space. Subsequent analysis uses adaptive Markov chain Monte Carlo methods and adaptive importance sampling to perform the integrals required for both inference (planet detection and orbit measurement), and information-maximizing sequential design (for adaptive scheduling of observations). We present an overview of our current techniques and highlight directions being explored by ongoing research.

astro-ph.IM↗

arXiv2018

Boosting Noise Robustness of Acoustic Model via Deep Adversarial Training

In realistic environments, speech is usually interfered by various noise and reverberation, which dramatically degrades the performance of automatic speech recognition (ASR) systems. To alleviate this issue, the commonest way is to use a well-designed speech enhancement approach as the front-end of ASR. However, more complex pipelines, more computations and even higher hardware costs (microphone array) are additionally consumed for this kind of methods. In addition, speech enhancement would result in speech distortions and mismatches to training. In this paper, we propose an adversarial training method to directly boost noise robustness of acoustic model. Specifically, a jointly compositional scheme of generative adversarial net (GAN) and neural network-based acoustic model (AM) is used in the training phase. GAN is used to generate clean feature representations from noisy features by the guidance of a discriminator that tries to distinguish between the true clean signals and generated signals. The joint optimization of generator, discriminator and AM concentrates the strengths of both GAN and AM for speech recognition. Systematic experiments on CHiME-4 show that the proposed method significantly improves the noise robustness of AM and achieves the average relative error rate reduction of 23.38% and 11.54% on the development and test set, respectively.

cs.SD↗