arXiv subjects

Feng Liu

Publications and source records attributed to Feng Liu.

At least 613 records · Page 34Linked to original sources

arXiv2019

Context-Aware Image Matting for Simultaneous Foreground and Alpha Estimation

Natural image matting is an important problem in computer vision and graphics. It is an ill-posed problem when only an input image is available without any external information. While the recent deep learning approaches have shown promising results, they only estimate the alpha matte. This paper presents a context-aware natural image matting method for simultaneous foreground and alpha matte estimation. Our method employs two encoder networks to extract essential information for matting. Particularly, we use a matting encoder to learn local features and a context encoder to obtain more global context information. We concatenate the outputs from these two encoders and feed them into decoder networks to simultaneously estimate the foreground and alpha matte. To train this whole deep neural network, we employ both the standard Laplacian loss and the feature loss: the former helps to achieve high numerical performance while the latter leads to more perceptually plausible results. We also report several data augmentation strategies that greatly improve the network's generalization performance. Our qualitative and quantitative experiments show that our method enables high-quality matting for a single natural image. Our inference codes and models have been made publicly available at https://github.com/hqqxyy/Context-Aware-Matting.

cs.CV↗

arXiv2019

IBRS: An Efficient Identity-based Batch Verification Scheme for VANETs Based on Ring Signature

Vehicular ad-hoc networks (VANETs) are one of the most important components in an Intelligent Transportation System (ITS), which aims to provide information communication between vehicles. A safety-critical vehicular communication requires security, privacy, auditability, and efficiency. To satisfy these requirements simultaneously, several conditional privacy-preserving authentication schemes are proposed by employing ring signature. However, these methods have been paid too little attention to the issues like \textit{how to choose the valid ring members} or \textit{how to set up a ring}. In this paper, we introduce an efficient conditional privacy-preserving scheme which provides an appropriate approach establishing the list of ring members. Moreover, our proposed scheme also supports batch verification to significantly reduce the computational cost. According to the analysis of security, our scheme is sufficiently resistant against several common attacks in VANETs. The performance results show that the proposed scheme is efficient and practical with both low computation and communication cost.

cs.CR↗

arXiv2019

Optical Control of transverse Motion of Ionization injected Electrons in Laser Plasma Wakefield

We demonstrate an all-optical method for controlling the transverse motion of ionization injected electron beam, by utilizing the transversely asymmetrical wakefield via adjusting the shape of the laser focal spot. When the laser spot shape is changed from the circular to the obliquely elliptic in experiment, the electron beam shape becomes from elliptic to three different shapes. The 3D-PIC simulation results agree well with experiment, and it shows the trajectories of the accelerated electrons change from undulation to helix. Such an all-optical method could be useful for convenient control of the transverse motion of an electron beam which results in synchrotron radiation with orbit angular momentum.

physics.plasm-ph↗

arXiv2019

3D Ken Burns Effect from a Single Image

The Ken Burns effect allows animating still images with a virtual camera scan and zoom. Adding parallax, which results in the 3D Ken Burns effect, enables significantly more compelling results. Creating such effects manually is time-consuming and demands sophisticated editing skills. Existing automatic methods, however, require multiple input images from varying viewpoints. In this paper, we introduce a framework that synthesizes the 3D Ken Burns effect from a single image, supporting both a fully automatic mode and an interactive mode with the user controlling the camera. Our framework first leverages a depth prediction pipeline, which estimates scene depth that is suitable for view synthesis tasks. To address the limitations of existing depth estimation methods such as geometric distortions, semantic distortions, and inaccurate depth boundaries, we develop a semantic-aware neural network for depth prediction, couple its estimate with a segmentation-based depth adjustment process, and employ a refinement neural network that facilitates accurate depth predictions at object boundaries. According to this depth estimate, our framework then maps the input image to a point cloud and synthesizes the resulting video frames by rendering the point cloud from the corresponding camera positions. To address disocclusions while maintaining geometrically and temporally coherent synthesis results, we utilize context-aware color- and depth-inpainting to fill in the missing information in the extreme views of the camera path, thus extending the scene geometry of the point cloud. Experiments with a wide variety of image content show that our method enables realistic synthesis results. Our study demonstrates that our system allows users to achieve better results while requiring little effort compared to existing solutions for the 3D Ken Burns effect creation.

cs.CV↗

arXiv2019

On Learning Disentangled Representations for Gait Recognition

Gait, the walking pattern of individuals, is one of the important biometrics modalities. Most of the existing gait recognition methods take silhouettes or articulated body models as gait features. These methods suffer from degraded recognition performance when handling confounding variables, such as clothing, carrying and viewing angle. To remedy this issue, we propose a novel AutoEncoder framework, GaitNet, to explicitly disentangle appearance, canonical and pose features from RGB imagery. The LSTM integrates pose features over time as a dynamic gait feature while canonical features are averaged as a static gait feature. Both of them are utilized as classification features. In addition, we collect a Frontal-View Gait (FVG) dataset to focus on gait recognition from frontal-view walking, which is a challenging problem since it contains minimal gait cues compared to other views. FVG also includes other important variations, e.g., walking speed, carrying, and clothing. With extensive experiments on CASIA-B, USF, and FVG datasets, our method demonstrates superior performance to the SOTA quantitatively, the ability of feature disentanglement qualitatively, and promising computational efficiency. We further compare our GaitNet with state-of-the-art face recognition to demonstrate the advantages of gait biometrics identification under certain scenarios, e.g., long distance/lower resolutions, cross viewing angles.

cs.CV↗

arXiv2019

Meta Learning with Relational Information for Short Sequences

This paper proposes a new meta-learning method -- named HARMLESS (HAwkes Relational Meta LEarning method for Short Sequences) for learning heterogeneous point process models from short event sequence data along with a relational network. Specifically, we propose a hierarchical Bayesian mixture Hawkes process model, which naturally incorporates the relational information among sequences into point process modeling. Compared with existing methods, our model can capture the underlying mixed-community patterns of the relational network, which simultaneously encourages knowledge sharing among sequences and facilitates adaptive learning for each individual sequence. We further propose an efficient stochastic variational meta expectation maximization algorithm that can scale to large problems. Numerical experiments on both synthetic and real data show that HARMLESS outperforms existing methods in terms of predicting the future events.

cs.LG↗

arXiv2019

Optical Black-hole Analog Created by Topological Phase Transition with a Long-lived Horizon

Hawking radiation, a manifestation of quantum field theory in curved spacetime, has stimulated extensive theoretical and experimental studies of various black-hole (BH) analogs. However, an undisputed confirmation of Hawking radiation remains elusive. One challenge is BH analog structures with long-lived horizons are difficult to achieve. Here, we theoretically demonstrate a new type of optical BH analog based on light cone evolution associated with topological phase transition of Dirac cones. The transition from a type-II to type-I Dirac/Weyl cone creates an analogous curved spacetime that crosses a type-III Dirac/Weyl cone, which affords a stationary configuration of long-lived event horizon. Photons tunneling through the horizon emit a spectrum of Hawking radiation. As an example, we design a laboratory version in an inhomogeneous two-dimensional graphyne-like topological photonic lattice with a Hawking temperature of 0.14 mK. Understanding Hawking-like radiation in this unique topological BH is not only of fundamental interest in its own right but may also provide new hints to gravitational physics.

cond-mat.mes-hall↗

arXiv2019

Asynchronous Distributed Voltage Control in Active Distribution Networks

With the explosion of distributed energy resources (DERs), voltage regulation in distribution networks has been facing a great challenge. This paper derives an asynchronous distributed voltage control strategy based on the partial primal-dual gradient algorithm, where both active and reactive controllable power of DERs are considered. Different types of asynchrony due to imperfect communication or practical limits, such as random time delays and non-identical sampling/control rates, are fitted into a unified analytic framework. The asynchronous algorithm is then converted into a fixed-point problem by employing the operator splitting method, which leads to a convergence proof with mild conditions. Moreover, an online implementation method is provided to make the controller adjustable to time-varying environments. Finally, numerical experiments are carried out on a rudimentary 8-bus system and the IEEE-123 distribution network to verify the effectiveness of the proposed method.

eess.SY↗

arXiv2019

MemeFaceGenerator: Adversarial Synthesis of Chinese Meme-face from Natural Sentences

Chinese meme-face is a special kind of internet subculture widely spread in Chinese Social Community Networks. It usually consists of a template image modified by some amusing details and a text caption. In this paper, we present MemeFaceGenerator, a Generative Adversarial Network with the attention module and template information as supplementary signals, to automatically generate meme-faces from text inputs. We also develop a web service as system demonstration of meme-face synthesis. MemeFaceGenerator has been shown to be capable of generating high-quality meme-faces from random text inputs.

cs.CL↗

arXiv2019

3D Face Modeling From Diverse Raw Scan Data

Traditional 3D face models learn a latent representation of faces using linear subspaces from limited scans of a single database. The main roadblock of building a large-scale face model from diverse 3D databases lies in the lack of dense correspondence among raw scans. To address these problems, this paper proposes an innovative framework to jointly learn a nonlinear face model from a diverse set of raw 3D scan databases and establish dense point-to-point correspondence among their scans. Specifically, by treating input scans as unorganized point clouds, we explore the use of PointNet architectures for converting point clouds to identity and expression feature representations, from which the decoder networks recover their 3D face shapes. Further, we propose a weakly supervised learning approach that does not require correspondence label for the scans. We demonstrate the superior dense correspondence and representation power of our proposed method, and its contribution to single-image 3D face reconstruction.

cs.CV↗

arXiv2019

Toward Distributed Stability Analytics for Power Systems with Heterogeneous Bus Dynamics

The stability issue emerges as a growing number of diverse power apparatus connecting to the power system. The stability analysis for such power systems is required to adapt to heterogeneity and scalability. This paper derives a local passivity index condition that guarantees the system-wide stability for lossless power systems with interconnected, nonlinear, heterogeneous bus dynamics. Our condition requires each bus dynamics to be output feedback passive with a large enough index w.r.t. a special supply rate. This condition fits for numerous existing models since it only constrains the input-output property rather than the detailed dynamics. Furthermore, for three typical examples of bus dynamics in power systems, we show that this condition can be reached via proper control designs. Simulations on a 3-bus heterogeneous power system verify our results in both lossless and lossy cases. The conservativeness of our condition is also demonstrated, as well as the impact on transient stability. It shows that our condition is quite tight and a larger index benefits transient stability.

eess.SY↗

arXiv2019

Towards Distributed Stability Analytics of Dynamic Power Systems: A Phasor-Circuit Theory Perspective

With the rapid development of renewable and distributed energies, the underlying dynamics of power systems are no longer dominated by large synchronous generators, but by numerous dynamic components with heterogeneous characteristics. In such a situation, the traditional stability analysis method may fail due to the challenges of heterogeneity and scalability. In this paper, we handle this issue by fundamental circuit theory. Inspired by the work of Brayton and Moser in the nonlinear RLC circuit, we extend the concept of the voltage potential to phasor circuits and offer new results into the distributed stability analytics in power systems. We show that under certain distributed passivity-like conditions the system-wide stability can be ensured. The simulation of a 3-bus system is also provided to verify our results.

eess.SY↗

arXiv2019

Data-driven subtraction of anisotropic flows in jet-like correlation studies in heavy-ion collisions

Measurements of two-particle azimuthal angle correlations are a useful tool to study the distribution of jet energy loss, however, they are complicated because of the significant anisotropic flow background. We devise a data-driven method for subtracting anisotropic flow background in jet-like correlation analysis. We first require a large recoil momentum ($P_x$) within a given pseudo-rapidity ($η$) range from a high-transverse momentum particle to enhance in-acceptance population of away-side jet-like correlations. Then we take the difference of two-particle correlations in the close-region and far-region with respect to the $η$ region of $P_x$ to subtract the anisotropic flow background. We use a toy model which contains only anisotropic flow and PYTHIA8 which have jets to demonstrate the validity of our data-driven method. The results indicate that the data-driven method can subtract anisotropic flow effectively.

nucl-ex↗

arXiv2019

Buy or Sell? Energy Sharing of Prosumers on Constrained Networks

The advent of intelligent agents who produce and consume energy by themselves has led the smart grid into the era of "prosumer", offering the energy system and customers a unique opportunity to revaluate/trade their spot energy via a sharing initiative. To this end, designing an appropriate sharing mechanism is an issue with crucial importance and has captured great attention. This paper addresses the prosumers' demand response problem via energy sharing. Under a general supply-demand function bidding scheme, a sharing market clearing procedure considering network constraints is proposed, which gives rise to a generalized Nash game. The existence and uniqueness of market equilibrium are proved in non-congested cases. When congestion occurs, infinitely much equilibrium may exist because the strategy spaces of prosumers are correlated. A price-regulation procedure is introduced in the sharing mechanism, which outcomes a unique equilibrium that is fair to all participants. Properties of the improved sharing mechanism, including the individual rational behaviors of prosumers and the components of sharing price, are revealed. When the number of prosumers increases, the proposed sharing mechanism approaches social optimum. Even with fixed number of resources, introducing competition can result in a decreasing social cost. Illustrative examples validate the theoretical results and provide more insights for the energy sharing research.

math.OC↗

arXiv2019

Dimensional Crossover of Topological Edge States in Su-Schrieffer-Heeger Model

Su-Schrieffer-Heeger (SSH) model on two-dimensional square lattice exhibits a topological phase transition, which is related to the Zak phase determined by bulk band topology. The strong modulation of electron hopping causes nontrivial charge polarization even in the presence of inversion symmetry. The energy band structures and topological edge states have been calculated numerically in previous studies. Here, however, full energy spectrum and explicit form of wave functions for two-dimensional bulk and one-dimensional ribbon geometries of SSH model are analytically derived using wave mechanics approach. Explicit analytic representations of wave functions provide the information of parity for each subband, localization length and critical point of topological phase transition in SSH ribbon. It is also shown that the dimensional crossover of topological transition point for SSH model from one to two-dimension.

cond-mat.mes-hall↗

arXiv2019

Probabilistic Structure Learning for EEG/MEG Source Imaging with Hierarchical Graph Prior

Brain source imaging is an important method for noninvasively characterizing brain activity using Electroencephalogram (EEG) or Magnetoencephalography (MEG) recordings. Traditional EEG/MEG Source Imaging (ESI) methods usually assume that either source activity at different time points is unrelated, or that similar spatiotemporal patterns exist across an entire study period. The former assumption makes ESI analyses sensitive to noise, while the latter renders ESI analyses unable to account for time-varying patterns of activity. To effectively deal with noise while maintaining flexibility and continuity among brain activation patterns, we propose a novel probabilistic ESI model based on a hierarchical graph prior. Under our method, a spanning tree constraint ensures that activity patterns have spatiotemporal continuity. An efficient algorithm based on alternating convex search is presented to solve the proposed model and is provably convergent. Comprehensive numerical studies using synthetic data on a real brain model are conducted under different levels of signal-to-noise ratio (SNR) from both sensor and source spaces. We also examine the EEG/MEG data in a real application, in which our ESI reconstructions are neurologically plausible. All the results demonstrate significant improvements of the proposed algorithm over the benchmark methods in terms of source localization performance, especially at high noise levels.

stat.AP↗

arXiv2019

Weyl points created by a three-dimensional flat band

Following the discovery of topological insulators (TIs), topological Dirac/Weyl semimetal has attracted much recent interest. A prevailing mechanism for the formation of Weyl points is by breaking time-reversal symmetry (TRS) or spatial inversion symmetry of a Dirac point. Here we demonstrate a generic formation mechanism for Weyl points by breaking TRS of a three-dimensional (3D) TI featured with highly degenerate 3D flat bands (FBs). It is in direct contrast to the conventional view that breaking TRS of a 2D/3D TI leads to a Chern insulator exhibiting quantum anomalous Hall effect. Based on a tight-binding model of pyrochlore lattice, we show that this unusual 3D-FB-enabled Weyl state may contain only a minimum of two Weyl points. Furthermore, using first-principles calculations, we identify this Weyl state in a real material Sn$_2$Nb$_2$O$_7$. The main features of the resulting Weyl points are analyzed with respect to symmetry, topological invariant and surface state. Our finding sheds new light on our fundamental understanding of topological physics and significantly extends the scope of Weyl semimetals to attract immediate experimental interest.

cond-mat.mes-hall↗

arXiv2019

Topological superconducting phase in high-Tc superconductor MgB2 with Dirac-nodal-line fermions

Topological superconductors are an intriguing and elusive quantum phase, characterized by topologically protected gapless surface/edge states residing in a bulk superconducting gap, which hosts Majorana fermions. Unfortunately, all currently known topological superconductors have a very low transition temperature, limiting experimental measurements of Majorana fermions. Here we discover the existence of a topological Dirac-nodal-line state in a well-known conventional high-temperature superconductor, MgB2. First-principles calculations show that the Dirac-nodal-line structure exhibits a unique one-dimensional dispersive Dirac-nodal line, protected by both spatial-inversion and time-reversal symmetry, which connects the electron and hole Dirac states. Most importantly, we show that the topological superconducting phase can be realized with a conventional s-wave superconducting gap, evidenced by the topological edge mode of the MgB2 thin films showing chiral edge states. Our discovery may enable the experimental measurement of Majorana fermions at high temperature.

cond-mat.supr-con↗