arXiv subjects

Hao Liu

Publications and source records attributed to Hao Liu.

At least 649 records · Page 36Linked to original sources

arXiv2018

Bees with attitude: the effect of gusts on flight dynamics

Flight is a complicated task at small scales in part due to the ubiquitous unsteady air which contains it. Flying organisms deal with these difficulties using active and passive control mechanisms to steer their body motion. Body attitudes of flapping organisms are linked with their resultant flight trajectories and performance, yet little is understood about how discrete unsteady aerodynamic phenomena affect the interlaced dynamics of such systems. In this study, we examined freely flying bumblebees subject to a single discrete gust to emulate aerodynamic disturbances encountered in nature. Bumblebees are expert commanders of the aerial domain as they persistently forage within complex terrain elements. Physical obstacles such as flowers produce local effects representative of a typified gust which threatens the precise control of intricate maneuvers. By tracking the 3D dynamics of bees flying through gusts, we determined the sequences of motion that permit flight in three disturbance conditions. Bees repetitively executed a series of passive impulsive maneuvers followed by active recovery maneuvers. Impulsive motion was unique in each gust direction, maintaining control purely by passive manipulation of the body. Bees pitched up and slowed-down at the beginning of recovery in every disturbance, followed by corrective maneuvers which brought attitudes back to their original state. Bees were displaced the most by the sideward gust, displaying large lateral translations and roll deviations. Upward gusts were easier for bees to fly through, causing only minor flight changes and minimal recovery times. Downward gusts severely impaired the control response of bees, inflicting strong adverse forces which sharply upset trajectories. Bees used interesting control strategies when flying in each disturbance, offering new insights into insect-scale flapping flight and bio-inspired robotic systems.

physics.flu-dyn↗

arXiv2018

A blind search for a common signal in gravitational wave detectors

We propose a blind, template-free method for the extraction of a common signal between the Hanford and Livingston detectors and apply it especially to the GW150914 event. We construct a log-likelihood method that maximizes the cross-correlation between each detector and the common signal and minimizes the cross-correlation between the residuals. The reliability of this method is tested using simulations with an injected common signal. Finally, our method is used to assess the quality of theoretical gravitational wave templates for GW150914.

astro-ph.IM↗

arXiv2018

Neutral and charged (pseduo)scalar mesons and diquarks under magnetic fields

We investigate both (pseudo)scalar mesons and diquarks in the presence of external magnetic field in the framework of the two-flavored Nambu--Jona-Lasinio (NJL) model, where mesons and diquarks are constructed by infinite sum of quark-loop chains by using random phase approximation. The polarization function of the quark-loop is calculated to the leading order of $1/N_c$ expansion by taking the quark propagator in the Landau level representation. We systematically investigate the masses behaviors of scalar $σ$ meson, neutral and charged pions as well as the scalar diquarks, with respect to the magnetic field strength at finite temperature and chemical potential. It is shown that the numerical results of both neutral and charged pions are consistent with the lattice QCD simulations. The mass of the charge neutral pion keeps almost a constant under the magnetic field, which is preserved by the remnant symmetry of QCD$\times$QED in the vacuum. The mass of the charge neutral scalar $σ$ is around two times quark mass and increases with the magnetic field due to the magnetic catalysis effect, which is an typical example showing that the polarized internal quark structure cannot be neglected when we consider the meson properties under magnetic field. For the charged particles, the one quark-antiquark loop contribution to the charged $π^{\pm}$ increases essentially with the increase of magnetic fields due to the magnetic catalysis of the polarized quarks. However, the one quark-quark loop contribution to the scalar diquark mass is negative comparing with the point-particle result and the loop effect is small.

hep-ph↗

arXiv2017

On Connecting Stochastic Gradient MCMC and Differential Privacy

Significant success has been realized recently on applying machine learning to real-world applications. There have also been corresponding concerns on the privacy of training data, which relates to data security and confidentiality issues. Differential privacy provides a principled and rigorous privacy guarantee on machine learning models. While it is common to design a model satisfying a required differential-privacy property by injecting noise, it is generally hard to balance the trade-off between privacy and utility. We show that stochastic gradient Markov chain Monte Carlo (SG-MCMC) -- a class of scalable Bayesian posterior sampling algorithms proposed recently -- satisfies strong differential privacy with carefully chosen step sizes. We develop theory on the performance of the proposed differentially-private SG-MCMC method. We conduct experiments to support our analysis and show that a standard SG-MCMC sampler without any modification (under a default setting) can reach state-of-the-art performance in terms of both privacy and utility on Bayesian learning.

stat.ML↗

arXiv2017

Dynamics, Structure and Glass Formation of Calcium Aluminate Liquids

Crystalline calcium-aluminates include the phases essential in the setting of Portland cements developed over the last century. It is only within recent decades, however, that calcium-aluminate melts and glasses have begun to attract attention, bringing new functionalities in photonic and electronic applications. These studies, though, have been limited to compositions close to a deep eutectic from where glasses easily form. With the development of contactless levitation furnaces the glass forming region can now be hugely extended. We have taken advantage of these developments to rationalise, for the first time, melt rheology with structural properties across this expanded compositional range, substantiating this with atomistic simulation. In the process, we have discovered that supercooled calcium-aluminates comprise a new system where fragile-to-strong phase transitions are ubiquitous. Taking this holistic approach, we have quantified the common basis of thermo-physical and structural diversity in this novel glass-forming system, together with its inherent polyamorphism.

cond-mat.mtrl-sci↗

arXiv2017

Interpretations of the DAMPE electron data

The DArk Matter Particle Explorer (DAMPE), a high energy cosmic ray and $γ$-ray detector in space, has recently reported the new measurement of the total electron plus positron flux between 25 GeV and 4.6 TeV. A spectral softening at $\sim0.9$ TeV and a tentative peak at $\sim1.4$ TeV have been reported. We study the physical implications of the DAMPE data in this work. The presence of the spectral break significantly tightens the constraints on the model parameters to explain the electron/positron excesses. The spectral softening can either be explained by the maximum acceleration limits of electrons by astrophysical sources, or a breakdown of the common assumption of continuous distribution of electron sources at TeV energies in space and time. The tentive peak at $\sim1.4$ TeV implies local sources of electrons/positrons with quasi-monochromatic injection spectrum. We find that the cold, ultra-relativistic $e^+e^-$ winds from pulsars may give rise to such a structure. The pulsar is requird to be middle-aged, relatively slowly-rotated, mildly magnetized, and isolated in a density cavity. The annihilation of DM particles ($m_χ\sim1.5$ TeV) into $e^+e^-$ pairs in a nearby clump or an over-density region may also explain the data. In the DM scenario, the inferred clump mass (or density enhancement) is about $10^7-10^8$ M$_\odot$ (or $17-35$ times of the canonical local density) assuming a thermal production cross section, which is relatively extreme compared with the expectation from numerical simulations. A moderate enhancement of the annihilation cross section via, e.g., the Sommerfeld mechanism or non-thermal production, is thus needed.

astro-ph.HE↗

arXiv2017

Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictions

Deep neural networks are widely used for classification. These deep models often suffer from a lack of interpretability -- they are particularly difficult to understand because of their non-linear nature. As a result, neural networks are often treated as "black box" models, and in the past, have been trained purely to optimize the accuracy of predictions. In this work, we create a novel network architecture for deep learning that naturally explains its own reasoning for each prediction. This architecture contains an autoencoder and a special prototype layer, where each unit of that layer stores a weight vector that resembles an encoded training input. The encoder of the autoencoder allows us to do comparisons within the latent space, while the decoder allows us to visualize the learned prototypes. The training objective has four terms: an accuracy term, a term that encourages every prototype to be similar to at least one encoded input, a term that encourages every encoded input to be close to at least one prototype, and a term that encourages faithful reconstruction by the autoencoder. The distances computed in the prototype layer are used as part of the classification process. Since the prototypes are learned during training, the learned network naturally comes with explanations for each prediction, and the explanations are loyal to what the network actually computes.

cs.AI↗

arXiv2017

Triangle Generative Adversarial Networks

A Triangle Generative Adversarial Network ($Δ$-GAN) is developed for semi-supervised cross-domain joint distribution matching, where the training data consists of samples from each domain, and supervision of domain correspondence is provided by only a few paired samples. $Δ$-GAN consists of four neural networks, two generators and two discriminators. The generators are designed to learn the two-way conditional distributions between the two domains, while the discriminators implicitly define a ternary discriminative function, which is trained to distinguish real data pairs and two kinds of fake data pairs. The generators and discriminators are trained together using adversarial learning. Under mild assumptions, in theory the joint distributions characterized by the two generators concentrate to the data distribution. In experiments, three different kinds of domain pairs are considered, image-label, image-image and image-attribute pairs. Experiments on semi-supervised image classification, image-to-image translation and attribute-based image generation demonstrate the superiority of the proposed approach.

cs.LG↗

arXiv2017

Optimal liquidation in a Level-I limit order book for large tick stocks

We propose a framework to study the optimal liquidation strategy in a limit order book for large-tick stocks, with spread equal to one tick. All order book events (market orders, limit orders and cancellations) occur according to independent Poisson processes, with parameters depending on price move directions. Our goal is to maximise the expected terminal wealth of an agent who needs to liquidate her positions within a fixed time horizon. Assuming that the agent trades (through sell limit order or/and sell market order) only when the price moves, we model her liquidation procedure as a semi-Markov decision process, and compute the semi-Markov kernel using Laplace method in the language of queueing theory. The optimal liquidation policy is then solved by dynamic programming, and illustrated numerically.

q-fin.TR↗

arXiv2017

ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching

We investigate the non-identifiability issues associated with bidirectional adversarial training for joint distribution matching. Within a framework of conditional entropy, we propose both adversarial and non-adversarial approaches to learn desirable matched joint distributions for unsupervised and supervised tasks. We unify a broad family of adversarial models as joint distribution matching problems. Our approach stabilizes learning of unsupervised bidirectional adversarial learning methods. Further, we introduce an extension for semi-supervised learning tasks. Theoretical results are validated in synthetic data and real-world applications.

stat.ML↗

arXiv2017

Downlink Energy Efficiency of Power Allocation and Wireless Backhaul Bandwidth Allocation in Heterogeneous Small Cell Networks

The widespread application of wireless services and dense devices access have triggered huge energy consumption. Because of the environmental and financial considerations, energy-efficient design in wireless networks becomes an inevitable trend. To the best of the authors' knowledge, energy-efficient orthogonal frequency division multiple access heterogeneous small cell optimization comprehensively considering energy efficiency maximization, power allocation, wireless backhaul bandwidth allocation, and user Quality of Service is a novel approach and research direction, and it has not been investigated. In this paper, we study the energy-efficient power allocation and wireless backhaul bandwidth allocation in orthogonal frequency division multiple access heterogeneous small cell networks. Different from the existing resource allocation schemes that maximize the throughput, the studied scheme maximizes energy efficiency by allocating both transmit power of each small cell base station to users and bandwidth for backhauling, according to the channel state information and the circuit power consumption. The problem is first formulated as a non-convex nonlinear programming problem and then it is decomposed into two convex subproblems. A near optimal iterative resource allocation algorithm is designed to solve the resource allocation problem. A suboptimal low-complexity approach is also developed by exploring the inherent structure and property of the energy-efficient design. Simulation results demonstrate the effectiveness of the proposed algorithms by comparing with the existing schemes.

cs.IT↗

arXiv2017

On the time lags of the LIGO signals

To date, the LIGO collaboration has detected three gravitational wave (GW) events appearing in both its Hanford and Livingston detectors. In this article we reexamine the LIGO data with regard to correlations between the two detectors. With special focus on GW150914, we report correlations in the detector noise which, at the time of the event, happen to be maximized for the same time lag as that found for the event itself. Specifically, we analyze correlations in the calibration lines in the vicinity of 35\,Hz as well as the residual noise in the data after subtraction of the best-fit theoretical templates. The residual noise for the other two events, GW151226 and GW170104, exhibits similar behavior. A clear distinction between signal and noise therefore remains to be established in order to determine the contribution of gravitational waves to the detected signals.

astro-ph.IM↗

arXiv2017

Neural Person Search Machines

We investigate the problem of person search in the wild in this work. Instead of comparing the query against all candidate regions generated in a query-blind manner, we propose to recursively shrink the search area from the whole image till achieving precise localization of the target person, by fully exploiting information from the query and contextual cues in every recursive search step. We develop the Neural Person Search Machines (NPSM) to implement such recursive localization for person search. Benefiting from its neural search mechanism, NPSM is able to selectively shrink its focus from a loose region to a tighter one containing the target automatically. In this process, NPSM employs an internal primitive memory component to memorize the query representation which modulates the attention and augments its robustness to other distracting regions. Evaluations on two benchmark datasets, CUHK-SYSU Person Search dataset and PRW dataset, have demonstrated that our method can outperform current state-of-the-arts in both mAP and top-1 evaluation protocols.

cs.CV↗

arXiv2017

Infinite families of 2-designs from GA_1(q) actions

Group action is a standard approach to obtain $t$-designs. In this approach, selecting a specific permutation group with a certain degree of transitivity or homogeneity and a proper set of base blocks is important for obtaining $t$-$(v, k, λ)$ designs with computable parameters $t, v, k$, and $λ$. The general affine group $\GA_1(q)$ is $2$-transitive on $\gf(q)$, and has relatively a small size. In this paper, we determine the parameters of a number of infinite families of $2$-designs obtained from the action of the group $\GA_1(q)$ on certain base blocks, and demonstrate that some of the $2$-designs give rise to linear codes with optimal or best parameters known. Open problems are also presented.

math.CO↗

arXiv2017

Video-based Person Re-identification with Accumulative Motion Context

Video based person re-identification plays a central role in realistic security and video surveillance. In this paper we propose a novel Accumulative Motion Context (AMOC) network for addressing this important problem, which effectively exploits the long-range motion context for robustly identifying the same person under challenging conditions. Given a video sequence of the same or different persons, the proposed AMOC network jointly learns appearance representation and motion context from a collection of adjacent frames using a two-stream convolutional architecture. Then AMOC accumulates clues from motion context by recurrent aggregation, allowing effective information flow among adjacent frames and capturing dynamic gist of the persons. The architecture of AMOC is end-to-end trainable and thus motion context can be adapted to complement appearance clues under unfavorable conditions (e.g. occlusions). Extensive experiments are conduced on three public benchmark datasets, i.e., the iLIDS-VID, PRID-2011 and MARS datasets, to investigate the performance of AMOC. The experimental results demonstrate that the proposed AMOC network outperforms state-of-the-arts for video-based re-identification significantly and confirm the advantage of exploiting long-range motion context for video based person re-identification, validating our motivation evidently.

cs.CV↗

arXiv2017

Stochastic Sequential Neural Networks with Structured Inference

Unsupervised structure learning in high-dimensional time series data has attracted a lot of research interests. For example, segmenting and labelling high dimensional time series can be helpful in behavior understanding and medical diagnosis. Recent advances in generative sequential modeling have suggested to combine recurrent neural networks with state space models (e.g., Hidden Markov Models). This combination can model not only the long term dependency in sequential data, but also the uncertainty included in the hidden states. Inheriting these advantages of stochastic neural sequential models, we propose a structured and stochastic sequential neural network, which models both the long-term dependencies via recurrent neural networks and the uncertainty in the segmentation and labels via discrete random variables. For accurate and efficient inference, we present a bi-directional inference network by reparamterizing the categorical segmentation and labels with the recent proposed Gumbel-Softmax approximation and resort to the Stochastic Gradient Variational Bayes. We evaluate the proposed model in a number of tasks, including speech modeling, automatic segmentation and labeling in behavior understanding, and sequential multi-objects recognition. Experimental results have demonstrated that our proposed model can achieve significant improvement over the state-of-the-art methods.

cs.LG↗

arXiv2017

Towards understanding the Planck thermal dust models

Understanding the properties of dust emission in the microwave domain is an important premise for the next generation of cosmic microwave background (CMB) experiments, devoted to the measurement of the primordial $B$-modes of polarization. In this paper, we compare three solutions to thermal dust emission by the Planck Collaboration \cite{PlanckDust03,planck_16,planck_com} to point out significant differences between their respective parameters (the spectral index $β$, the optical depth $τ$ and the dust temperature $T_d$). These differences originate from e.g. the priors on the parameters or the contribution of the Cosmic infrared background (CIB).

astro-ph.CO↗

arXiv2017

End-to-End Comparative Attention Networks for Person Re-identification

Person re-identification across disjoint camera views has been widely applied in video surveillance yet it is still a challenging problem. One of the major challenges lies in the lack of spatial and temporal cues, which makes it difficult to deal with large variations of lighting conditions, viewing angles, body poses and occlusions. Recently, several deep learning based person re-identification approaches have been proposed and achieved remarkable performance. However, most of those approaches extract discriminative features from the whole frame at one glimpse without differentiating various parts of the persons to identify. It is essentially important to examine multiple highly discriminative local regions of the person images in details through multiple glimpses for dealing with the large appearance variance. In this paper, we propose a new soft attention based model, i.e., the end to-end Comparative Attention Network (CAN), specifically tailored for the task of person re-identification. The end-to-end CAN learns to selectively focus on parts of pairs of person images after taking a few glimpses of them and adaptively comparing their appearance. The CAN model is able to learn which parts of images are relevant for discerning persons and automatically integrates information from different parts to determine whether a pair of images belongs to the same person. In other words, our proposed CAN model simulates the human perception process to verify whether two images are from the same person. Extensive experiments on three benchmark person re-identification datasets, including CUHK01, CHUHK03 and Market-1501, clearly demonstrate that our proposed end-to-end CAN for person re-identification outperforms well established baselines significantly and offer new state-of-the-art performance.

cs.CV↗