arXiv subjects

Hui Li

Publications and source records attributed to Hui Li.

At least 559 records · Page 31Linked to original sources

arXiv2022

Research Opportunities in Plasma Astrophysics

Major scientific questions and research opportunities are described on 10 unprioritized plasma astrophysics topics: (1) magnetic reconnection, (2) collisionless shocks and particle acceleration, (3) waves and turbulence, (4) magnetic dynamos, (5) interface and shear instabilities, (6) angular momentum transport, (7) dusty plasmas, (8) radiative hydrodynamics, (9) relativistic, pair-dominated and strongly magnetized plasmas, (10) jets and outflows. Note that this is a conference report from a Workshop on Opportunities in Plasma Astrophysics (WOPA, https://w3.pppl.gov/conferences/2010/WOPA/) in January 2010, that attracted broad representation from the community and was supported by the U.S. Department of Energy, National Aeronautics and Space Administration, National Science Foundation, American Physical Society's Topical Group for Plasma Astrophysics and Division of Plasma Physics, and Center for Magnetic Self-Organization in Laboratory and Astrophysical Plasmas. Although there has been much planning and many developments in both science and infrastructure since the report was written, most of the motivation, priorities, problems and technical challenges discussed therein remain unaddressed and are relevant at the time of posting.

physics.plasm-ph↗

arXiv2022

Infrared and visible image fusion using Latent Low-Rank Representation

Infrared and visible image fusion is an important problem in the field of image fusion which has been applied widely in many fields. To better preserve the useful information from source images, in this paper, we propose a novel image fusion method based on latent low-rank representation(LatLRR) which is simple and effective. Firstly, the source images are decomposed into low-rank parts(global structure) and salient parts(local structure) by LatLRR. Then, the low-rank parts are fused by weighted-average strategy to preserve more contour information. Then, the salient parts are simply fused by sum strategy which is a efficient operation in this fusion framework. Finally, the fused image is obtained by combining the fused low-rank part and the fused salient part. Compared with other fusion methods experimentally, the proposed method has better fusion performance than state-of-the-art fusion methods in both subjective and objective evaluation. The Code of our fusion method is available at https://github.com/hli1221/imagefusion\_Infrared\_visible\_latlrr

cs.CV↗

arXiv2022

Multi-focus Noisy Image Fusion using Low-Rank Representation

Multi-focus noisy image fusion represents an important task in the field of image fusion which generates a single, clear and focused image from all source images. In this paper, we propose a novel multi-focus noisy image fusion method based on low-rank representation (LRR) which is a powerful tool in representation learning. A multi-scale transform framework is adopted in which source images are decomposed into low frequency and high frequency coefficients, respectively. For low frequency coefficients, the fused low frequency coefficients are determined by a spatial frequency strategy, while the high frequency coefficients are fused by the LRR-based fusion strategy. Finally, the fused image is reconstructed by inverse multi-scale transforms with fused coefficients. Experimental results demonstrate that the proposed algorithm offers state-of-the-art performance even when the source images contain noise. The Code of our fusion method is available at https://github.com/hli1221/imagefusion_noisy_lrr

cs.CV↗

arXiv2022

Multi-focus Image Fusion Based on Similarity Characteristics

A novel multi-focus image fusion algorithm performed in spatial domain based on similarity characteristics is proposed incorporating with region segmentation. In this paper, a new similarity measure is developed based on the structural similarity (SSIM) index, which is more suitable for multi-focus image segmentation. Firstly, the SSNSIM map is calculated between two input images. Then we segment the SSNSIM map using watershed method, and merge the small homogeneous regions with fuzzy c-means clustering algorithm (FCM). For three source images, a joint region segmentation method based on segmentation of two images is used to obtain the final segmentation result. Finally, the corresponding segmented regions of the source images are fused according to their average gradient. The performance of the image fusion method is evaluated by several criteria including spatial frequency, average gradient, entropy, edge retention etc. The evaluation results indicate that the proposed method is effective and has good visual perception.

cs.CV↗

arXiv2022

Improved dual channel pulse coupled neural network and its application to multi-focus image fusion

This paper presents an improved dual channel pulse coupled neural network (IDC-PCNN) model for image fusion. The model can overcome some defects of standard PCNN model. In this fusion scheme, the multiplication rule is replaced by addition rule in the information fusion pool of dual channel PCNN (DC-PCNN) model. Meanwhile the sum of modified Laplacian (SML) measure is adopted, which is better than other focus measures. This method not only inherits the good characteristics of the standard PCNN model but also enhances the computing efficiency and fusion quality. The performance of the proposed method is evaluated by using four criteria including average cross entropy, root mean square error, peak value signal to noise ratio and structure similarity index. Comparative studies show that the proposed fusion algorithm outperforms the standard PCNN method and the DC-PCNN method.

cs.CV↗

arXiv2022

Variational Policy Propagation for Multi-agent Reinforcement Learning

We propose a \emph{collaborative} multi-agent reinforcement learning algorithm named variational policy propagation (VPP) to learn a \emph{joint} policy through the interactions over agents. We prove that the joint policy is a Markov Random Field under some mild conditions, which in turn reduces the policy space effectively. We integrate the variational inference as special differentiable layers in policy such that the actions can be efficiently sampled from the Markov Random Field and the overall policy is differentiable. We evaluate our algorithm on several large scale challenging tasks and demonstrate that it outperforms previous state-of-the-arts.

cs.LG↗

arXiv2022

Femtosecond rotational dynamics of D$_2$ molecules in superfluid helium nanodroplets

Rotational dynamics of D$_2$ molecules inside helium nanodroplets is induced by a moderately intense femtosecond (fs) pump pulse and measured as a function of time by recording the yield of HeD$^+$ ions, created through strong-field dissociative ionization with a delayed fs probe pulse. The yield oscillates with a period of 185 fs, reflecting field-free rotational wave packet dynamics, and the oscillation persists for more than 500 periods. Within the experimental uncertainty, the rotational constant BHe of the in-droplet D$_2$ molecule, determined by Fourier analysis, is the same as Bgas for an isolated D$_2$ molecule. Our observations show that the D$_2$ molecules inside helium nanodroplets essentially rotate as free D$_2$ molecules.

physics.atm-clus↗

arXiv2022

Exploring Fusion Strategies for Accurate RGBT Visual Object Tracking

We address the problem of multi-modal object tracking in video and explore various options of fusing the complementary information conveyed by the visible (RGB) and thermal infrared (TIR) modalities including pixel-level, feature-level and decision-level fusion. Specifically, different from the existing methods, paradigm of image fusion task is heeded for fusion at pixel level. Feature-level fusion is fulfilled by attention mechanism with channels excited optionally. Besides, at decision level, a novel fusion strategy is put forward since an effortless averaging configuration has shown the superiority. The effectiveness of the proposed decision-level fusion strategy owes to a number of innovative contributions, including a dynamic weighting of the RGB and TIR contributions and a linear template update operation. A variant of which produced the winning tracker at the Visual Object Tracking Challenge 2020 (VOT-RGBT2020). The concurrent exploration of innovative pixel- and feature-level fusion strategies highlights the advantages of the proposed decision-level fusion method. Extensive experimental results on three challenging datasets, \textit{i.e.}, GTOT, VOT-RGBT2019, and VOT-RGBT2020, demonstrate the effectiveness and robustness of the proposed method, compared to the state-of-the-art approaches. Code will be shared at \textcolor{blue}{\emph{https://github.com/Zhangyong-Tang/DFAT}.

cs.CV↗

arXiv2022

On the Existence of Fast Modes in Compressible Magnetohydrodynamic Turbulence

We study the existence and property of Fast magnetosonic modes in 3D compressible MHD turbulence by carrying out a number of simulations with compressible and incompressible driving conditions. We use two approaches to determine the presence of Fast modes: mode decomposition based on spatial variations only and spatio-temporal 4D-FFT analysis of all fluctuations. The latter method enables us to quantify fluctuations that satisfy the dispersion relation of Fast modes with finite frequency. Overall, we find that the fraction of Fast modes identified via spatio-temporal 4D FFT approach in total fluctuation power is either tiny with nearly incompressible driving or ~2% with highly compressible driving. We discuss the implications of our results for understanding the compressible fluctuations in space and astrophysics plasmas.

physics.space-ph↗

arXiv2022

Global spin polarization of multistrange hyperons and feed-down effect in heavy-ion collisions

Global spin polarization of hyperons is an important observable to probe the vorticity of the quark-gluon plasma produced in heavy-ion collisions. We calculate the global polarizations of $Λ$, $Ξ^-$, and $Ω^-$ in Au+Au collisions at energies $\sqrt{s_\text{NN}}=$ 7.7--200 GeV based on a multiphase transport model. Our calculations suggest that their primary global polarizations fulfill $P_{Ω^-}\simeq 5/3P_{Ξ^-}\simeq 5/3P_Λ$. We also estimate the feed-down effect of particle decay on the global polarization. With the feed-down effect taken into account, the global polarizations of $Λ$ and $Ξ^-$ are clearly separated, and the final global polarizations are in the ordering: $P_{Ω^-}>P_{Ξ^-}>P_Λ$. Such a relation can be tested in experiments, which will provide us more information about the global polarization mechanism.

nucl-th↗

arXiv2022

Switchable and unswitchable bulk photovoltaic effect in two-dimensional interlayer-sliding ferroelectrics

Spontaneous polarization and bulk photovoltaic effect (BPVE) are two concomitant physical properties in ferroelectric materials. The flipping of ferroelectric order usually accompanies with the switching of BPVE as both of them are reversed under the inversion symmetry. In this study, we report the distinctive BPVE characters in two-dimensional (2D) interlayer sliding ferroelectric materials featuring unswitchable in-plane BPVE (light-induced photocurrent in the xy plane) and switchable out-of-plane BPVE (light-induced polarization along the z-direction). Symmetry analysis within abstract bilayer crystal model and first-principles calculations validate these BPVE properties. It is because the positive and negative ferroelectric states caused by interlayer sliding are related by mirror symmetry which cannot flip all the BPVE tensor elements. This finding extends the understanding of the relationship between ferroelectricity and BPVE. On one hand, the switchable out-of-plane BPVE can be used to design switchable photoelectric devices. On the other hand, the in-plane BPVE is robust against the ferroelectric flipping, and the unswitchable character is beneficial to construct larger-scale photoelectric devices.

cond-mat.mtrl-sci↗

arXiv2021

Magnetic Energy Conversion in MHD: Curvature Relaxation and Perpendicular Expansion of Magnetic Fields

The mechanisms and pathways of magnetic energy conversion are an important subject for many laboratory, space and astrophysical systems. Here, we present a perspective on magnetic energy conversion in MHD through magnetic field curvature relaxation (CR) and perpendicular expansion (PE) due to magnetic pressure gradients, and quantify their relative importance in two representative cases, namely 3D magnetic reconnection and 3D kink-driven instability in an astrophysical jet. We find that the CR and PE processes have different temporal and spatial evolution in these systems. The relative importance of the two processes tends to reverse as the system enters the nonlinear stage from the instability growth stage. Overall, the two processes make comparable contributions to the magnetic energy conversion with the PE process somewhat stronger than the CR process. We further explore how these energy conversion terms can be related to particle energization in these systems.

physics.space-ph↗

arXiv2021

Prospects for assembling ultracold radioactive molecules from laser-cooled atoms

Molecules with unstable isotopes often contain heavy and deformed nuclei and thus possess a high sensitivity to parity-violating effects, such as Schiff moments. Currently the best limits on Schiff moments are set with diamagnetic atoms. Polar molecules with quantum-enhanced sensing capabilities, however, can offer better sensitivity. In this work, we consider the prototypical 223Fr107Ag molecule, as the octupole deformation of the unstable 223Fr francium nucleus amplifies the nuclear Schiff moment of the molecule by two orders of magnitude relative to that of spherical nuclei and as the silver atom has a large electronegativity. To develop a competitive experimental platform based on molecular quantum systems, 223Fr atoms and 107Ag atoms have to be brought together at ultracold temperatures. That is, we explore the prospects of forming 223Fr107Ag from laser-cooled Fr and Ag atoms. We have performed fully relativistic electronic-structure calculations of ground and excited states of FrAg that account for the strong spin-dependent relativistic effects of Fr and the strong ionic bond to Ag. In addition, we predict the nearest-neighbor densities of magnetic-field Feshbach resonances in ultracold 223Fr+107Ag collisions with coupled-channel calculations. These resonances can be used for magneto-association into ultracold, weakly-bound FrAg. We also determine the conditions for creating 223Fr107Ag molecules in their absolute ground state from these weakly-bound dimers via stimulated Raman adiabatic passage using our calculations of the relativistic transition electronic dipole moments.

quant-ph↗

arXiv2021

Structured Multimodal Attentions for TextVQA

In this paper, we propose an end-to-end structured multimodal attention (SMA) neural network to mainly solve the first two issues above. SMA first uses a structural graph representation to encode the object-object, object-text and text-text relationships appearing in the image, and then designs a multimodal graph attention network to reason over it. Finally, the outputs from the above modules are processed by a global-local attentional answering module to produce an answer splicing together tokens from both OCR and general vocabulary iteratively by following M4C. Our proposed model outperforms the SoTA models on TextVQA dataset and two tasks of ST-VQA dataset among all models except pre-training based TAP. Demonstrating strong reasoning ability, it also won first place in TextVQA Challenge 2020. We extensively test different OCR methods on several reasoning models and investigate the impact of gradually increased OCR performance on TextVQA benchmark. With better OCR results, different models share dramatic improvement over the VQA accuracy, but our model benefits most blessed by strong textual-visual reasoning ability. To grant our method an upper bound and make a fair testing base available for further works, we also provide human-annotated ground-truth OCR annotations for the TextVQA dataset, which were not given in the original release. The code and ground-truth OCR annotations for the TextVQA dataset are available at https://github.com/ChenyuGAO-CS/SMA

cs.CV↗

arXiv2021

Enhanced countering adversarial attacks via input denoising and feature restoring

Despite the fact that deep neural networks (DNNs) have achieved prominent performance in various applications, it is well known that DNNs are vulnerable to adversarial examples/samples (AEs) with imperceptible perturbations in clean/original samples. To overcome the weakness of the existing defense methods against adversarial attacks, which damages the information on the original samples, leading to the decrease of the target classifier accuracy, this paper presents an enhanced countering adversarial attack method IDFR (via Input Denoising and Feature Restoring). The proposed IDFR is made up of an enhanced input denoiser (ID) and a hidden lossy feature restorer (FR) based on the convex hull optimization. Extensive experiments conducted on benchmark datasets show that the proposed IDFR outperforms the various state-of-the-art defense methods, and is highly effective for protecting target models against various adversarial black-box or white-box attacks. \footnote{Souce code is released at: \href{https://github.com/ID-FR/IDFR}{https://github.com/ID-FR/IDFR}}

cs.CV↗

arXiv2021

Directly Characterizing the Coherence of Quantum Detectors by Sequential Measurement

The quantum properties of quantum measurements are indispensable resources in quantum information processing and have drawn extensive research interest. The conventional approach to reveal the quantum properties relies on the reconstruction of the entire measurement operators by quantum detector tomography. However, many specific properties can be determined by a part of matrix entries of the measurement operators, which provides us the possibility to simplify the process of property characterization. Here, we propose a general framework to directly obtain individual matrix entries of the measurement operators by sequentially measuring two non-compatible observables. This method allows us to circumvent the complete tomography of the quantum measurement and extract the useful information for our purpose. We experimentally implement this scheme to monitor the coherent evolution of a general quantum measurement by determining the off-diagonal matrix entries. The investigation of the measurement precision indicates the good feasibility of our protocol to the arbitrary quantum measurements. Our results pave the way for revealing the quantum properties of quantum measurements by selectively determining the matrix entries of the measurement operators.

quant-ph↗

arXiv2021

A novel multiobjective evolutionary algorithm based on decomposition and multi-reference points strategy

Many real-world optimization problems such as engineering design can be eventually modeled as the corresponding multiobjective optimization problems (MOPs) which must be solved to obtain approximate Pareto optimal fronts. Multiobjective evolutionary algorithm based on decomposition (MOEA/D) has been regarded as a significantly promising approach for solving MOPs. Recent studies have shown that MOEA/D with uniform weight vectors is well-suited to MOPs with regular Pareto optimal fronts, but its performance in terms of diversity usually deteriorates when solving MOPs with irregular Pareto optimal fronts. In this way, the solution set obtained by the algorithm can not provide more reasonable choices for decision makers. In order to efficiently overcome this drawback, we propose an improved MOEA/D algorithm by virtue of the well-known Pascoletti-Serafini scalarization method and a new strategy of multi-reference points. Specifically, this strategy consists of the setting and adaptation of reference points generated by the techniques of equidistant partition and projection. For performance assessment, the proposed algorithm is compared with existing four state-of-the-art multiobjective evolutionary algorithms on benchmark test problems with various types of Pareto optimal fronts. According to the experimental results, the proposed algorithm exhibits better diversity performance than that of the other compared algorithms. Finally, our algorithm is applied to two real-world MOPs in engineering optimization successfully.

cs.NE↗

arXiv2021

The $\sin2ϕ_h$ azimuthal asymmetry of pion production in SIDIS within TMD factorization

We study the $\sin2ϕ_h$ azimuthal asymmetry of charged and neutral pion productions off the longitudinally polarized nucleon targets in semi-inclusive deeply inelastic scattering (SIDIS) process within the transverse momentum dependent~(TMD) factorization. The asymmetry is contributed by the convolution of the TMD distribution $h_{1L}^{\perp}$ and the Collins fragmentation function $H_{1}^\perp$. We adopt the Wandzura-Wilczek-type~(WW-type) approximation for $h_{1L}^{\perp}$ and two different parametrizations on the nonperturbative part of TMD evolution for $h_{1L}^{\perp}$ and $H_{1}^\perp$. We numerically estimate the asymmetry in the kinematical configurations available at HERMES, CLAS and CLAS12. It is shown that our theoretical calculation can describe the HERMES and CLAS data, except the asymmetry of the $π^-$ production off proton target at HERMES and CLAS. We also find that different choices of the nonperturbative part of TMD evolution lead qualitatively similar results.

hep-ph↗