arXiv subjects

Jingbo Wang

Publications and source records attributed to Jingbo Wang.

At least 181 records · Page 10Linked to original sources

arXiv2020

Wideband Monitoring Observations of PSR J1803-3002A in the Globular Cluster NGC 6522

We report the first wideband monitoring observations of PSR J1803-3002A, a relatively bright millisecond pulsar in the globular cluster NGC 6522 with a spin period of 7.1 ms and no known binary companion. These observations were performed using the Parkes 64-m radio telescope with the Ultra-Wideband Low (UWL) receiver system, which covers 704 to 4032 MHz. We confirm that PSR J1803-3002A is an isolated millisecond pulsar located near the cluster center and probe the emission properties of the pulsar over the wide observed band. The mean pulse profile consists of three components, with the outer components becoming more prominent at higher frequencies, and a mean spectral index for the pulsed emission of -1.66+/-0.07 over the observed band. The fractional linear and circular polarization increase with increasing frequency, which is unusual for pulsars. We determine a Faraday rotation measure of -107+/-6 rad m-2 for the pulsar. PSR J1803-3002A is a distant pulsar in the Galactic plane, but there is no evidence of pulse broadening due to interstellar scattering in our observations. These results demonstrate the power of ultra-wideband receiving and signal processing systems.

astro-ph.HE↗

arXiv2020

Fractional charge of quasi-particles on the horizon of black holes

In the previous works, it was claimed that black holes can be considered as topological insulators. In this paper, we will show that they are actually fractional topological insulators. That is, the quasi-particles and quasi-holes can have fractional charges and statistics (spins). For BTZ black hole, the filling fraction is $v=\frac{1}{2k}$. For Kerr black hole, the filling fraction is $v=\fracπ{S}$, where $S$ is the entropy of black hole.

gr-qc↗

arXiv2020

NeuroDiff: Scalable Differential Verification of Neural Networks using Fine-Grained Approximation

As neural networks make their way into safety-critical systems, where misbehavior can lead to catastrophes, there is a growing interest in certifying the equivalence of two structurally similar neural networks. For example, compression techniques are often used in practice for deploying trained neural networks on computationally- and energy-constrained devices, which raises the question of how faithfully the compressed network mimics the original network. Unfortunately, existing methods either focus on verifying a single network or rely on loose approximations to prove the equivalence of two networks. Due to overly conservative approximation, differential verification lacks scalability in terms of both accuracy and computational cost. To overcome these problems, we propose NeuroDiff, a symbolic and fine-grained approximation technique that drastically increases the accuracy of differential verification while achieving many orders-of-magnitude speedup. NeuroDiff has two key contributions. The first one is new convex approximations that more accurately bound the difference neurons of two networks under all possible inputs. The second one is judicious use of symbolic variables to represent neurons whose difference bounds have accumulated significant error. We also find that these two techniques are complementary, i.e., when combined, the benefit is greater than the sum of their individual benefits. We have evaluated NeuroDiff on a variety of differential verification tasks. Our results show that NeuroDiff is up to 1000X faster and 5X more accurate than the state-of-the-art tool.

cs.LG↗

arXiv2020

Testing the weak equivalence principle with the binary neutron star merger GW170817: the gravitational contribution of the host galaxy

The successful detection of the binary neutron star (BNS) merger GW170817 and its electromagnetic (EM) counterparts has provided an opportunity to explore the joint effect of the host galaxy and the Milky Way (MW) on the weak equivalence principle (WEP) test. In this paper, using the Navarro$-$Frenk$-$White (NFW) profile and the Herquist profile, we present an analytic model to calculate the galactic potential, in which the possible locations of the source by the observed angle offset and the second supernova (SN2) kick are accounted for. We show that the upper limit of $Δγ$ is $10^{-9}$ for the comparison between GW170817 and a gamma-ray burst (GRB 170817A), and it is $10^{-4}$ for the comparison between GW170817 and a bright optical transient (SSS17a, now with the IAU identification of AT 2017gfo). These limits are more stringent by one to two orders of magnitude than those determined solely by the measured MW potential in the literature. We demonstrate that the WEP test is strengthened by contribution from the host galaxy to the Shapiro time delay. Meanwhile, we also find that large natal kicks produce a maximum deviation of about $20\%$ to the results with a typical kick velocity 400$\sim$ 500 km s$^{-1}$. Finally, we analyze the impact from the halo mass of NGC 4993 with a typical 0.2 dex uncertainty, and find that the upper limit of $Δγ$, with a maximum mass $10^{12.4}h^{-1} M_{\odot}$, is nearly two times more stringent than that of the minimum mass $10^{12.0}h^{-1} M_{\odot}$.

astro-ph.HE↗

arXiv2020

Malleable 2.5D Convolution: Learning Receptive Fields along the Depth-axis for RGB-D Scene Parsing

Depth data provide geometric information that can bring progress in RGB-D scene parsing tasks. Several recent works propose RGB-D convolution operators that construct receptive fields along the depth-axis to handle 3D neighborhood relations between pixels. However, these methods pre-define depth receptive fields by hyperparameters, making them rely on parameter selection. In this paper, we propose a novel operator called malleable 2.5D convolution to learn the receptive field along the depth-axis. A malleable 2.5D convolution has one or more 2D convolution kernels. Our method assigns each pixel to one of the kernels or none of them according to their relative depth differences, and the assigning process is formulated as a differentiable form so that it can be learnt by gradient descent. The proposed operator runs on standard 2D feature maps and can be seamlessly incorporated into pre-trained CNNs. We conduct extensive experiments on two challenging RGB-D semantic segmentation dataset NYUDv2 and Cityscapes to validate the effectiveness and the generalization ability of our method.

cs.CV↗

arXiv2020

Bi-directional Cross-Modality Feature Propagation with Separation-and-Aggregation Gate for RGB-D Semantic Segmentation

Depth information has proven to be a useful cue in the semantic segmentation of RGB-D images for providing a geometric counterpart to the RGB representation. Most existing works simply assume that depth measurements are accurate and well-aligned with the RGB pixels and models the problem as a cross-modal feature fusion to obtain better feature representations to achieve more accurate segmentation. This, however, may not lead to satisfactory results as actual depth data are generally noisy, which might worsen the accuracy as the networks go deeper. In this paper, we propose a unified and efficient Cross-modality Guided Encoder to not only effectively recalibrate RGB feature responses, but also to distill accurate depth information via multiple stages and aggregate the two recalibrated representations alternatively. The key of the proposed architecture is a novel Separation-and-Aggregation Gating operation that jointly filters and recalibrates both representations before cross-modality aggregation. Meanwhile, a Bi-direction Multi-step Propagation strategy is introduced, on the one hand, to help to propagate and fuse information between the two modalities, and on the other hand, to preserve their specificity along the long-term propagation process. Besides, our proposed encoder can be easily injected into the previous encoder-decoder structures to boost their performance on RGB-D semantic segmentation. Our model outperforms state-of-the-arts consistently on both in-door and out-door challenging datasets. Code of this work is available at https://charlescxk.github.io/

cs.CV↗

arXiv2020

QSW_MPI: a framework for parallel simulation of quantum stochastic walks

QSW_MPI is a python package developed for time-series simulation of continuous-time quantum stochastic walks. This model allows for the study of Markovian open quantum systems in the Lindblad formalism, including a generalisation of the continuous-time random walk and continuous-time quantum walk. Consisting of a python interface accessing parallelised Fortran libraries utilising sparse data structures, QSW_MPI is scalable to massively parallel computers, which makes possible the simulation of a wide range of walk dynamics on directed and undirected graphs of arbitrary complexity.

quant-ph↗

arXiv2020

Motion Guided 3D Pose Estimation from Videos

We propose a new loss function, called motion loss, for the problem of monocular 3D Human pose estimation from 2D pose. In computing motion loss, a simple yet effective representation for keypoint motion, called pairwise motion encoding, is introduced. We design a new graph convolutional network architecture, U-shaped GCN (UGCN). It captures both short-term and long-term motion information to fully leverage the additional supervision from the motion loss. We experiment training UGCN with the motion loss on two large scale benchmarks: Human3.6M and MPI-INF-3DHP. Our model surpasses other state-of-the-art models by a large margin. It also demonstrates strong capacity in producing smooth 3D sequences and recovering keypoint motion.

cs.CV↗

arXiv2020

Combinatorial optimisation via highly efficient quantum walks

We present a highly efficient quantum circuit for performing continuous time quantum walks (CTQWs) over an exponentially large set of combinatorial objects, provided that the objects can be indexed efficiently. CTQWs form the core mixing operation of a generalised version of the Quantum Approximate Optimisation Algorithm, which works by `steering' the quantum amplitude into high-quality solutions. The efficient quantum circuit holds the promise of finding high-quality solutions to certain classes of NP-hard combinatorial problems such as the Travelling Salesman Problem, maximum set splitting, graph partitioning, and lattice path optimisation.

quant-ph↗

arXiv2020

BiSeNet V2: Bilateral Network with Guided Aggregation for Real-time Semantic Segmentation

The low-level details and high-level semantics are both essential to the semantic segmentation task. However, to speed up the model inference, current approaches almost always sacrifice the low-level details, which leads to a considerable accuracy decrease. We propose to treat these spatial details and categorical semantics separately to achieve high accuracy and high efficiency for realtime semantic segmentation. To this end, we propose an efficient and effective architecture with a good trade-off between speed and accuracy, termed Bilateral Segmentation Network (BiSeNet V2). This architecture involves: (i) a Detail Branch, with wide channels and shallow layers to capture low-level details and generate high-resolution feature representation; (ii) a Semantic Branch, with narrow channels and deep layers to obtain high-level semantic context. The Semantic Branch is lightweight due to reducing the channel capacity and a fast-downsampling strategy. Furthermore, we design a Guided Aggregation Layer to enhance mutual connections and fuse both types of feature representation. Besides, a booster training strategy is designed to improve the segmentation performance without any extra inference cost. Extensive quantitative and qualitative evaluations demonstrate that the proposed architecture performs favourably against a few state-of-the-art real-time semantic segmentation approaches. Specifically, for a 2,048x1,024 input, we achieve 72.6% Mean IoU on the Cityscapes test set with a speed of 156 FPS on one NVIDIA GeForce GTX 1080 Ti card, which is significantly faster than existing methods, yet we achieve better segmentation accuracy.

cs.CV↗

arXiv2020

Context Prior for Scene Segmentation

Recent works have widely explored the contextual dependencies to achieve more accurate segmentation results. However, most approaches rarely distinguish different types of contextual dependencies, which may pollute the scene understanding. In this work, we directly supervise the feature aggregation to distinguish the intra-class and inter-class context clearly. Specifically, we develop a Context Prior with the supervision of the Affinity Loss. Given an input image and corresponding ground truth, Affinity Loss constructs an ideal affinity map to supervise the learning of Context Prior. The learned Context Prior extracts the pixels belonging to the same category, while the reversed prior focuses on the pixels of different classes. Embedded into a conventional deep CNN, the proposed Context Prior Layer can selectively capture the intra-class and inter-class contextual dependencies, leading to robust feature representation. To validate the effectiveness, we design an effective Context Prior Network (CPNet). Extensive quantitative and qualitative evaluations demonstrate that the proposed model performs favorably against state-of-the-art semantic segmentation approaches. More specifically, our algorithm achieves 46.3% mIoU on ADE20K, 53.9% mIoU on PASCAL-Context, and 81.3% mIoU on Cityscapes. Code is available at https://git.io/ContextPrior.

cs.CV↗

arXiv2020

Hawking Radiation Is Pure

Information loss paradox is still a challenging problem in theoretical physics. In this essay, for statics BTZ black holes and Schwarzschild black holes, we propose a simple solution based on the similarity between black holes and topological insulators. That is, the Hawking radiation is pure due to the entanglement between the left-moving sector and right-moving sector of the Hawking radiation. And this entanglement may be detected in an analogue black hole in the near future.

gr-qc↗

arXiv2020

Quantum approximate algorithm for NP optimization problems with constraints

The Quantum Approximate Optimization Algorithm (QAOA) is an algorithmic framework for finding approximate solutions to combinatorial optimization problems, derived from an approximation to the Quantum Adiabatic Algorithm (QAA). In solving combinatorial optimization problems with constraints in the context of QAOA or QAA, one needs to find a way to encode problem constraints into the scheme. In this paper, we formalize different constraint types to linear equalities, linear inequalities, and arbitrary form. Based on this, we propose constraint-encoding schemes well-fitting into the QAOA framework for solving NP combinatorial optimization problems. The implemented algorithms demonstrate the effectiveness and efficiency of the proposed scheme by the testing results of varied instances of some well-known NP optimization problems. We argue that our work leads to a generalized framework for finding, in the context of QAOA, high-quality approximate solutions to combinatorial problems with various types of constraints.

quant-ph↗

arXiv2020

ReluDiff: Differential Verification of Deep Neural Networks

As deep neural networks are increasingly being deployed in practice, their efficiency has become an important issue. While there are compression techniques for reducing the network's size, energy consumption and computational requirement, they only demonstrate empirically that there is no loss of accuracy, but lack formal guarantees of the compressed network, e.g., in the presence of adversarial examples. Existing verification techniques such as Reluplex, ReluVal, and DeepPoly provide formal guarantees, but they are designed for analyzing a single network instead of the relationship between two networks. To fill the gap, we develop a new method for differential verification of two closely related networks. Our method consists of a fast but approximate forward interval analysis pass followed by a backward pass that iteratively refines the approximation until the desired property is verified. We have two main innovations. During the forward pass, we exploit structural and behavioral similarities of the two networks to more accurately bound the difference between the output neurons of the two networks. Then in the backward pass, we leverage the gradient differences to more accurately compute the most beneficial refinement. Our experiments show that, compared to state-of-the-art verification tools, our method can achieve orders-of-magnitude speedup and prove many more properties than existing tools.

cs.LG↗

arXiv2020

Microscopic states of Kerr black holes from boundary-bulk correspondence

It was claimed by the author that black holes can be considered as topological insulators. They both have boundary modes and those boundary modes can be described by an effective BF theory. In this paper, we analyze the boundary modes on the horizon of black holes with the methods developed for topological insulator. Firstly the BTZ black hole is analysed, and the results are compatible with the previous works. Then we generalize those results to Kerr black holes. Some new results are obtained: dimensionless right- and left-temperature can be defined and have well behaviors both in Schwarzschild limit $a\rightarrow 0$ and extremal limit $a\rightarrow M$. Upon the Kerr/CFT correspondence, we can associate a central charge $c=12 M r_+$ with an arbitrary Kerr black hole if a dual CFT exists. We can identify the microstates of the Kerr black hole with the quantum states of this scalar field. From this identification we can count the number of microstates of the Kerr black hole and give the Bekenstein-Hawking area law for the entropy.

gr-qc↗

arXiv2020

Central charges for Kerr and Kerr-AdS black holes in diverse dimensions

In the previous work we give a microscopic explanation of the entropy for the BTZ black hole and four-dimensional Kerr black hole based on the massless scalar field theory on the horizon. An essential input is the central charges of those black holes. In this paper, we calculate the central charges for Kerr black holes and Kerr-AdS black holes in diverse dimensions by rewriting the entropy formula in a suggesting way. Then we also give the statistical explanation for the entropy of those black holes based on the scalar field on the horizon which similar to 4D kerr black hole.

hep-th↗

arXiv2019

Hawking Radiation from Boundary Scalar Field

Hawking radiation is one essential property of quantum black hole. It results in the information loss paradox, and give important clue to the unification of quantum mechanics and general relativity. In the previous works, the boundary scalar field on the horizon of black holes were used to give the microstates of BTZ black holes and Kerr black holes. They are account for the Bekenstein-Hawking entropy. In this paper, we show that the Hawking radiation can also be derived from those scalar fields. Actually the Hawking radiation are superposition of thermal radiation of right/left sector at different temperatures. We also discuss the impact on the information loss paradox.

gr-qc↗

arXiv2019

Parkes Pulsar Timing Array constraints on ultralight scalar-field dark matter

It is widely accepted that dark matter contributes about a quarter of the critical mass-energy density in our Universe. The nature of dark matter is currently unknown, with the mass of possible constituents spanning nearly one hundred orders of magnitude. The ultralight scalar field dark matter, consisting of extremely light bosons with $m \sim 10^{-22}$ eV and often called "fuzzy" dark matter, provides intriguing solutions to some challenges at sub-Galactic scales for the standard cold dark matter model. As shown by Khmelnitsky and Rubakov, such a scalar field in the Galaxy would produce an oscillating gravitational potential with nanohertz frequencies, resulting in periodic variations in the times of arrival of radio pulses from pulsars. The Parkes Pulsar Timing Array (PPTA) has been monitoring 20 millisecond pulsars at two to three weeks intervals for more than a decade. In addition to the detection of nanohertz gravitational waves, PPTA offers the opportunity for direct searches for fuzzy dark matter in an astrophysically feasible range of masses. We analyze the latest PPTA data set which includes timing observations for 26 pulsars made between 2004 and 2016. We perform a search in this data set for evidence of ultralight dark matter in the Galaxy using Bayesian and Frequentist methods. No statistically significant detection has been made. We therefore place upper limits on the local dark matter density. Our limits, improving on previous searches by a factor of two to five, constrain the dark matter density of ultralight bosons with $m \leq 10^{-23}$ eV to be below $6\,\text{GeV}\,\text{cm}^{-3}$ with 95\% confidence in the Earth neighborhood. Finally, we discuss the prospect of probing the astrophysically favored mass range $m \gtrsim 10^{-22}$ eV with next-generation pulsar timing facilities.

astro-ph.CO↗