arXiv subjects

Hao Liu

Publications and source records attributed to Hao Liu.

At least 667 records · Page 37Linked to original sources

arXiv2017

All binary linear codes that are invariant under $\PSL_2(n)$

The projective special linear group $\PSL_2(n)$ is $2$-transitive for all primes $n$ and $3$-homogeneous for $n \equiv 3 \pmod{4}$ on the set $\{0,1, \cdots, n-1, \infty\}$. It is known that the extended odd-like quadratic residue codes are invariant under $\PSL_2(n)$. Hence, the extended quadratic residue codes hold an infinite family of $2$-designs for primes $n \equiv 1 \pmod{4}$, an infinite family of $3$-designs for primes $n \equiv 3 \pmod{4}$. To construct more $t$-designs with $t \in \{2, 3\}$, one would search for other extended cyclic codes over finite fields that are invariant under the action of $\PSL_2(n)$. The objective of this paper is to prove that the extended quadratic residue binary codes are the only nontrivial extended binary cyclic codes that are invariant under $\PSL_2(n)$.

cs.IT↗

arXiv2017

Parameters of two classes of LCD BCH codes

Historically, LCD cyclic codes were referred to as reversible cyclic codes, which had application in data storage. Due to a newly discovered application in cryptography, there has been renewed interest on LCD codes. In this paper, we explore two special families of LCD cyclic codes, which are both BCH codes. The dimensions and the minimum distances of these LCD BCH codes are investigated. As a byproduct, the parameters of some primitive BCH codes are also obtained.

cs.IT↗

arXiv2017

Learning Policies for Markov Decision Processes from Data

We consider the problem of learning a policy for a Markov decision process consistent with data captured on the state-actions pairs followed by the policy. We assume that the policy belongs to a class of parameterized policies which are defined using features associated with the state-action pairs. The features are known a priori, however, only an unknown subset of them could be relevant. The policy parameters that correspond to an observed target policy are recovered using $\ell_1$-regularized logistic regression that best fits the observed state-action samples. We establish bounds on the difference between the average reward of the estimated and the original policy (regret) in terms of the generalization error and the ergodic coefficient of the underlying Markov chain. To that end, we combine sample complexity theory and sensitivity analysis of the stationary distribution of Markov chains. Our analysis suggests that to achieve regret within order $O(\sqrtε)$, it suffices to use training sample size on the order of $Ω(\log n \cdot poly(1/ε))$, where $n$ is the number of the features. We demonstrate the effectiveness of our method on a synthetic robot navigation example.

math.OC↗

arXiv2016

Recurrent Image Captioner: Describing Images with Spatial-Invariant Transformation and Attention Filtering

Along with the prosperity of recurrent neural network in modelling sequential data and the power of attention mechanism in automatically identify salient information, image captioning, a.k.a., image description, has been remarkably advanced in recent years. Nonetheless, most existing paradigms may suffer from the deficiency of invariance to images with different scaling, rotation, etc.; and effective integration of standalone attention to form a holistic end-to-end system. In this paper, we propose a novel image captioning architecture, termed Recurrent Image Captioner (\textbf{RIC}), which allows visual encoder and language decoder to coherently cooperate in a recurrent manner. Specifically, we first equip CNN-based visual encoder with a differentiable layer to enable spatially invariant transformation of visual signals. Moreover, we deploy an attention filter module (differentiable) between encoder and decoder to dynamically determine salient visual parts. We also employ bidirectional LSTM to preprocess sentences for generating better textual representations. Besides, we propose to exploit variational inference to optimize the whole architecture. Extensive experimental results on three benchmark datasets (i.e., Flickr8k, Flickr30k and MS COCO) demonstrate the superiority of our proposed architecture as compared to most of the state-of-the-art methods.

cs.CV↗

arXiv2016

Panoptic Studio: A Massively Multiview System for Social Interaction Capture

We present an approach to capture the 3D motion of a group of people engaged in a social interaction. The core challenges in capturing social interactions are: (1) occlusion is functional and frequent; (2) subtle motion needs to be measured over a space large enough to host a social group; (3) human appearance and configuration variation is immense; and (4) attaching markers to the body may prime the nature of interactions. The Panoptic Studio is a system organized around the thesis that social interactions should be measured through the integration of perceptual analyses over a large variety of view points. We present a modularized system designed around this principle, consisting of integrated structural, hardware, and software innovations. The system takes, as input, 480 synchronized video streams of multiple people engaged in social activities, and produces, as output, the labeled time-varying 3D structure of anatomical landmarks on individuals in the space. Our algorithm is designed to fuse the "weak" perceptual processes in the large number of views by progressively generating skeletal proposals from low-level appearance cues, and a framework for temporal refinement is also presented by associating body parts to reconstructed dense 3D trajectory stream. Our system and method are the first in reconstructing full body motion of more than five people engaged in social interactions without using markers. We also empirically demonstrate the impact of the number of views in achieving this goal.

cs.CV↗

arXiv2016

City traffic forecasting using taxi GPS data: A coarse-grained cellular automata model

City traffic is a dynamic system of enormous complexity. Modeling and predicting city traffic flow remains to be a challenge task and the main difficulties are how to specify the supply and demands and how to parameterize the model. In this paper we attempt to solve these problems with the help of large amount of floating car data. We propose a coarse-grained cellular automata model that simulates vehicles moving on uniform grids whose size are much larger compared with the microscopic cellular automata model. The car-car interaction in the microscopic model is replaced by the coupling between vehicles and coarse-grained state variables in our model. To parameterize the model, flux-occupancy relations are fitted from the historical data at every grids, which serve as the coarse-grained fundamental diagrams coupling the occupancy and speed. To evaluate the model, we feed it with the historical travel demands and trajectories obtained from the floating car data and use the model to predict road speed one hour into the future. Numerical results show that our model can capture the traffic flow pattern of the entire city and make reasonable predictions. The current work can be considered a prototype for a model-based forecasting system for city traffic.

cs.OH↗

arXiv2016

Added costs of insect-scale flapping flight in unsteady airflows

The aerial environment in the operating domain of small-scale natural and artificial flapping wing fliers is highly complex, unsteady and generally turbulent. Considering flapping flight in an unsteady wind environment with a periodically varying lateral velocity component, we show that body rotations experienced by flapping wing fliers result in the reorientation of the aerodynamic force vector that can render a substantial cumulative deficit in the vertical force. We derive quantitative estimates of the body roll amplitude and the related energetic requirements to maintain the weight support in free flight under such conditions. We conduct force measurements of a miniature hummingbird-inspired robotic flapper and numerical simulations of a bumblebee. In both cases, we demonstrate the loss of weight support due to body roll rotations. Using semi-restrained flight measurements, we demonstrate the increased power requirements to maintain altitude in unsteady winds, achieved by increasing the flapping frequency. Flapping fliers may increase their flapping frequency as well as the stroke amplitude to produce the required increase in aerodynamic force, both of these two types of compensatory control requiring additional energetic cost. We analyze the existing data from experiments on animals flying in von Kármán streets and find reasonable agreement with the proposed theoretical model.

physics.flu-dyn↗

arXiv2016

Possible associated signal with GW150914 in the LIGO data

We present a simple method for the identification of weak signals associated with gravitational wave events. Its application reveals a signal with the same time lag as the GW150914 event in the released LIGO strain data with a significance around $3.2σ$. This signal starts about 10 minutes before GW150914 and lasts for about 45 minutes. Subsequent tests suggest that this signal is likely to be due to external sources.

astro-ph.IM↗

arXiv2016

Molecular Dynamics Simulation of Strong Shock Waves Propagating in Dense Deuterium With the Effect of Excited Electrons

We present a molecular dynamics simulation of shock waves propagating in dense deuterium with the electron force field method [J. T. Su and W. A. Goddard, Phys. Rev. Lett. 99, 185003 (2007)], which explicitly takes the excitation of electrons into consideration. Non-equilibrium features associated with the excitation of electrons are systematically investigated. We show that chemical bonds in D$_2$ molecules lead to a more complicated shock wave structure near the shock front, compared with the results of classical molecular dynamics simulation. Charge separation can bring about accumulation of net charges on the large scale, instead of the formation of a localized dipole layer, which might cause extra energy for the shock wave to propagate. In addition, the simulations also display that molecular dissociation at the shock front is the major factor corresponding to the "bump" structure in the principal Hugoniot. These results could help to build a more realistic picture of shock wave propagation in fuel materials commonly used in the inertial confinement fusion.

physics.plasm-ph↗

arXiv2016

Dimensions of three types of BCH codes over GF(q)

BCH codes have been studied for over fifty years and widely employed in consumer devices, communication systems, and data storage systems. However, the dimension of BCH codes is settled only for a very small number of cases. In this paper, we study the dimensions of BCH codes over finite fields with three types of lengths $n$, namely $n=q^m-1$, $n=(q^m-1)/(q-1)$ and $n=q^m+1$. For narrow-sense primitive BCH codes with designed distance $δ$, we investigate their dimensions for $δ$ in the range $1\le δ\le q^{\lceil\frac{m}{2}\rceil+1}$. For non-narrow sense primitive BCH codes, we provide two general formulas on their dimensions and give the dimensions explicitly in some cases. Furthermore, we settle the minimum distances of some primitive BCH codes. We also explore the dimensions of the BCH codes of lengths $n=(q^m-1)/(q-1)$ and $n=q^m+1$ over finite fields.

cs.IT↗

arXiv2016

Understanding the LIGO GW150914 event

We present a simplified method for the extraction of meaningful signals from Hanford and Livingstone 32 seconds data for the GW150914 event made publicly available by the LIGO collaboration and demonstrate its ability to reproduce the LIGO collaboration's own results quantitatively given the assumption that all narrow peaks in the power spectrum are a consequence of physically uninteresting signals and can be removed. After the clipping of these peaks and return to the time domain, the GW150914 event is readily distinguished from broadband background noise. This simple technique allows us to identify the GW150914 event without any assumption regarding its physical origin and with minimal assumptions regarding its shape. We also confirm that the LIGO GW150914 event is uniquely correlated in the Hanford and Livingston detectors for 4096 second data at the level of $6-7\,σ$ with a temporal displacement of $τ=6.9 \pm 0.4\,$ms. We have also identified a few events that are morphologically close to GW150914 but less strongly cross correlated with it.

astro-ph.CO↗

arXiv2016

A Family of Reversible BCH Codes

Cyclic codes are an interesting class of linear codes due to their efficient encoding and decoding algorithms as well as their theoretical importance. BCH codes form a subclass of cyclic codes and are very important in both theory and practice as they have good error-correcting capability and are widely used in communication systems, storage devices and consumer electronics. However, the dimension and minimum distance of BCH codes are not known in general. The objective of this paper is to study the dimension and minimum distance of a family of BCH codes over finite fields, i.e., a class of reversible BCH codes.

cs.IT↗

arXiv2016

Critical points of the cosmic velocity field and the uncertainties in the value of the Hubble constant

The existence of critical points for the peculiar velocity field is a natural feature of the correlated vector field. These points appear at the junctions of velocity domains with different orientations of their averaged velocity vectors. Since peculiar velocities are the important cause of the scatter in the Hubble expansion rate, we propose that a more precise determination of the Hubble constant can be made by restricting analysis to a subsample of observational data containing only the zones around the critical points of the peculiar velocity field, associated with voids and saddle points. On large-scales the critical points, where the first derivative of the gravitational potential vanishes, can easily be identified using the density field and classified by the behavior of the Hessian of the gravitational potential. We use high-resolution N-body simulations to show that these regions are stable in time and hence are excellent tracers of the initial conditions. Furthermore, we show that the variance of the Hubble flow can be substantially minimized by restricting observations to the subsample of such regions of vanishing velocity instead of aiming at increasing the statistics by averaging indiscriminately using the full data sets, as is the common approach.

astro-ph.CO↗

arXiv2016

Possible formation of high temperature superconductor at early stage of heavy-ion collisions

We investigate the effect of the inverse magnetic catalysis (IMC) on the charged $ρ$ meson condensation at finite temperature in the framework of the Nambu--Jona-Lasinio model, where mesons are calculated to the leading order of $1/N_c$ expansion by summing up infinity quark-loops. IMC for chiral condensate has been considered in three different ways, i.e. fitting Lattice data, using the running coupling constant and introducing the chiral chemical potential, respectively. It is observed that, with no IMC effect included, the critical magnetic field $eB_c$ for charged $ρ$ condensation increases monotonically with the temperature. However, including IMC substantially affects the polarized charged $ρ$ condensation around the critical temperature $T_c$ of chiral phase transition, the critical magnetic field $eB_c$ for charged $ρ$ condensation decreases with the temperature firstly, reaches to a minimum value around $T_c$, then increases with the temperature. Our calculation indicates that the charged $ρ$ condensation can exist in the temperature region of $1-1.5 T_c$ with critical magnetic field $eB_c\sim 0.15-0.3 {\rm GeV}^2$, which suggests that high temperature superconductor might be created through non-central heavy ion collisions at LHC energies. We also show that a growing electric conductivity in early stage of non-central heavy-ion collisions substantially delays the decay of strong magnetic field, which is helpful for the formation of the high temperature superconductor.

hep-ph↗

arXiv2016

Footprints of Loop I on Cosmic Microwave Background Maps

Cosmology has made enormous progress through studies of the cosmic microwave background, however the subtle signals being now sought such as B-mode polarisation due to primordial gravitational waves are increasingly hard to disentangle from residual Galactic foregrounds in the derived CMB maps. We revisit our finding that on large angular scales there are traces of the nearby old supernova remnant Loop I in the WMAP 9-year map of the CMB and confirm this with the new SMICA map from the Planck satellite.

astro-ph.CO↗

arXiv2016

Shear induced solidification of athermal systems with weak attraction

We find that unjammed packings of frictionless particles with rather weak attraction can always be driven into solid-like states by shear. The structure of shear-driven solids evolves continuously with packing fraction from gel-like to jamming-like, but is almost independent of the shear stress. In contrast, both the density of vibrational states (DOVS) and force network evolve progressively with the shear stress. There exists a packing fraction independent shear stress $σ_c$, at which the shear-driven solids are isostatic and have a flattened DOVS. Solid-like states induced by a shear stress greater than $σ_c$ possess properties of marginally jammed solids and are thus strictly-defined shear jammed states. Below $σ_c$, states at all packing fractions are under isostaticity and share common features in the DOVS and force network, although their structures can be rather different. Our study reveals the significance of the shear stress in determining properties of shear-driven solids and leads to an enriched jamming phase diagram for weakly attractive particles.

cond-mat.soft↗

arXiv2016

Bumblebees minimize control challenges by combining active and passive modes in unsteady winds

The natural wind environment that volant insects encounter is unsteady and highly complex, posing significant flight control and stability challenges. Unsteady airflows can range from structured chains of discrete vortices shed in the wake of an object to fully developed chaotic turbulence. It is critical to understand the flight control strategies insects employ to safely navigate in natural environments. We combined experiments on free flying bumblebees with high fidelity numerical simulations and lower order modeling to identify the salient mechanics that mediate insect flight in unsteady winds. We trained bumblebees to fly upwind towards an artificial flower in a wind tunnel under steady wind and in a von Karman street (23Hz) formed in the wake of a cylinder. The bees displayed significantly higher movement in the unsteady vortex street compared to steady winds. Correlation analysis revealed that at lower frequencies, less than 10 Hz, in both steady and unsteady winds the bees mediated lateral movement with body roll, typical casting motion. At higher frequencies in unsteady winds there was a negative correlation between body roll and lateral accelerations. Numerical simulations of a bumblebee in similar conditions permitted the separation of the passive and active components of the flight trajectories. Comparison between the free-flying and numerical bees revealed a novel mechanism that enables bees to passively ride out high frequency perturbations while performing active maneuvers and corrections at lower frequencies. The capacity of maintaining stability by combining passive and active modes at different timescales provides a viable means for volant animals and machines to tackle the control challenges posed by complex airflows.

physics.bio-ph↗

arXiv2016

Aerodynamic ground effect in fruitfly sized insect takeoff

Aerodynamic ground effect in flapping-wing insect flight is of importance to comparative morphologies and of interest to the micro-air-vehicle (MAV) community. Recent studies, however, show apparently contradictory results of either some significant extra lift or power savings, or zero ground effect. Here we present a numerical study of fruitfly sized insect takeoff with a specific focus on the significance of leg thrust and wing kinematics. Flapping-wing takeoff is studied using numerical modelling and high performance computing. The aerodynamic forces are calculated using a three-dimensional Navier--Stokes solver based on a pseudo-spectral method with volume penalization. It is coupled with a flight dynamics solver that accounts for the body weight, inertia and the leg thrust, while only having two degrees of freedom: the vertical and the longitudinal horizontal displacement. The natural voluntary takeoff of a fruitfly is considered as reference. The parameters of the model are then varied to explore possible effects of interaction between the flapping-wing model and the ground plane. These modified takeoffs include cases with decreased leg thrust parameter, and/or with periodic wing kinematics, constant body pitch angle. The results show that the ground effect during natural voluntary takeoff is negligible. In the modified takeoffs, when the rate of climb is slow, the difference in the aerodynamic forces due to the interaction with the ground is up to 6%. Surprisingly, depending on the kinematics, the difference is either positive or negative, in contrast to the intuition based on the helicopter theory, which suggests positive excess lift. This effect is attributed to unsteady wing-wake interactions. A similar effect is found during hovering.

physics.flu-dyn↗