arXiv subjects

Yi Jiang

Publications and source records attributed to Yi Jiang.

At least 199 records · Page 11Linked to original sources

arXiv2022

Ptychopy: GPU framework for ptychographic data analysis

X-ray ptychography imaging at synchrotron facilities like the Advanced Photon Source (APS) involves controlling instrument hardwares to collect a set of diffraction patterns from overlapping coherent illumination spots on extended samples, managing data storage, reconstructing ptychographic images from acquired diffraction patterns, and providing the visualization of results and feedback. In addition to the complicated workflow, ptychography instrument could produce up to several TB's of data per second that is needed to be processed in real time. This brings up the need to develop a high performance, robust and user friendly processing software package for ptychographic data analysis. In this paper we present a software framework which provides functionality of visualization, work flow control, and data reconstruction. To accelerate the computation and large datasets process, the data reconstruction part is implemented with three algorithms, ePIE, DM and LSQML using CUDA-C on GPU.

physics.ins-det↗

arXiv2022

A Direct Parallel-in-Time Quasi-Boundary Value Method for Inverse Space-Dependent Source Problems

Inverse source problems arise often in real-world applications, such as localizing unknown groundwater contaminant sources. Being different from Tikhonov regularization, the quasi-boundary value method has been proposed and analyzed as an effective way for regularizing such inverse source problems, which was shown to achieve an optimal order convergence rate under suitable assumptions. However, fast direct or iterative solvers for the resulting all-at-once large-scale linear systems have been rarely studied in the literature. In this work, we first proposed and analyzed a modified quasi-boundary value method, and then developed a diagonalization-based parallel-in-time (PinT) direct solver, which can achieve a dramatic speedup in CPU times when compared with MATLAB's sparse direct solver. In particular, the time-discretization matrix $B$ is shown to be diagonalizable, and the condition number of its eigenvector matrix $V$ is proven to exhibit quadratic growth, which guarantees the roundoff errors due to diagonalization is well controlled. Several 1D and 2D examples are presented to demonstrate the very promising computational efficiency of our proposed method, where the CPU times in 2D cases can be speedup by three orders of magnitude.

math.NA↗

arXiv2021

Sensitivity studies on the CKM angle $γ$ in $Λ_b^0 \to DΛ$ decays

The sensitivity of the CKM angle $γ$ in $Λ_b^0 \to D Λ$ decays has been studied using the decay parameter $α$ as an observable in addition to the decay rate asymmetry. Feasibility studies show that adding this observable improves the sensitivity on $γ$ by up to 60\% and makes the decays one of the most promising places to measure angle $γ$ and to search for $C\!P$ violation in $b$-baryon decays.

hep-ph↗

arXiv2021

Two-loop master integrals for the single top production associated with $W$ boson

The $tW$ associated production has the second-largest cross section among three single top production channels. A complete study of NNLO QCD corrections to $tW$ production is still missing in literature. We present the calculation of part of the requisite two-loop master integrals for the process $b+g \to t+W$ at NNLO QCD. It turns out that the 80 master integrals in two families, involving up to three massive internal lines, can be expressed by Goncharov polylogarithms. The canonical differential equations that these integrals obey contain at most three square roots. By separating the canonical master integrals into groups according to their dependences on roots, we succeed in achieving optimization of final expressions, which are validated by independent numerical checks.

hep-ph↗

arXiv2021

Numerical Study of Non-uniqueness for 2D Compressible Isentropic Euler Equations

In this paper, we numerically study a class of solutions with spiraling singularities in vorticity for two-dimensional, inviscid, compressible Euler systems, where the initial data have an algebraic singularity in vorticity at the origin. These are different from the multi-dimensional Riemann problems widely studied in the literature. Our computations provide numerical evidence of the existence of initial value problems with multiple solutions, thus revealing a fundamental obstruction toward the well-posedness of the governing equations. The compressible Euler equations are solved using the positivity-preserving discontinuous Galerkin method.

math.AP↗

arXiv2021

Quantitatively Predicting Modal Thermal Conductivity of Nanocrystalline Si by full band Monte Carlo simulations

Thermal transport of nanocrystalline Si is of great importance for the application of thermoelectrics. A better understanding of the modal thermal conductivity of nanocrystalline Si will be expected to benefit the efficiency of thermoelectrics. In this work, the variance reduced Monte Carlo simulation with full band of phonon dispersion is applied to study the modal thermal conductivity of nanocrystalline Si. Importantly, the phonon modal transmissions across the grain boundaries which are modeled by the amorphous Si interface are calculated by the mode-resolved atomistic Greens function method. The predicted ratios of thermal conductivity of nanocrystalline Si to that of bulk Si agree well with that of the experimental measurements in a wide range of grain size. The thermal conductivity of nanocrystalline Si is decreased from 54 percent to 3 percent and the contribution of phonons with mean free path larger than the grain size increases from 30 percent to 96 percnet as the grain size decreases from 550 nm to 10 nm. This work demonstrates that the full band Monte Carlo simulation using phonon modal transmission by the mode-resolved atomistic Greens function method can capture the phonon transport picture in complex nanostructures, and therefore can provide guidance for designing high performance Si based thermoelectrics.

cond-mat.mtrl-sci↗

arXiv2021

What Makes for End-to-End Object Detection?

Object detection has recently achieved a breakthrough for removing the last one non-differentiable component in the pipeline, Non-Maximum Suppression (NMS), and building up an end-to-end system. However, what makes for its one-to-one prediction has not been well understood. In this paper, we first point out that one-to-one positive sample assignment is the key factor, while, one-to-many assignment in previous detectors causes redundant predictions in inference. Second, we surprisingly find that even training with one-to-one assignment, previous detectors still produce redundant predictions. We identify that classification cost in matching cost is the main ingredient: (1) previous detectors only consider location cost, (2) by additionally introducing classification cost, previous detectors immediately produce one-to-one prediction during inference. We introduce the concept of score gap to explore the effect of matching cost. Classification cost enlarges the score gap by choosing positive samples as those of highest score in the training iteration and reducing noisy positive samples brought by only location cost. Finally, we demonstrate the advantages of end-to-end object detection on crowded scenes. The code is available at: \url{https://github.com/PeizeSun/OneNet}.

cs.CV↗

arXiv2021

TransTrack: Multiple Object Tracking with Transformer

In this work, we propose TransTrack, a simple but efficient scheme to solve the multiple object tracking problems. TransTrack leverages the transformer architecture, which is an attention-based query-key mechanism. It applies object features from the previous frame as a query of the current frame and introduces a set of learned object queries to enable detecting new-coming objects. It builds up a novel joint-detection-and-tracking paradigm by accomplishing object detection and object association in a single shot, simplifying complicated multi-step settings in tracking-by-detection methods. On MOT17 and MOT20 benchmark, TransTrack achieves 74.5\% and 64.5\% MOTA, respectively, competitive to the state-of-the-art methods. We expect TransTrack to provide a novel perspective for multiple object tracking. The code is available at: \url{https://github.com/PeizeSun/TransTrack}.

cs.CV↗

arXiv2021

Sparse R-CNN: End-to-End Object Detection with Learnable Proposals

We present Sparse R-CNN, a purely sparse method for object detection in images. Existing works on object detection heavily rely on dense object candidates, such as $k$ anchor boxes pre-defined on all grids of image feature map of size $H\times W$. In our method, however, a fixed sparse set of learned object proposals, total length of $N$, are provided to object recognition head to perform classification and location. By eliminating $HWk$ (up to hundreds of thousands) hand-designed object candidates to $N$ (e.g. 100) learnable proposals, Sparse R-CNN completely avoids all efforts related to object candidates design and many-to-one label assignment. More importantly, final predictions are directly output without non-maximum suppression post-procedure. Sparse R-CNN demonstrates accuracy, run-time and training convergence performance on par with the well-established detector baselines on the challenging COCO dataset, e.g., achieving 45.0 AP in standard $3\times$ training schedule and running at 22 fps using ResNet-50 FPN model. We hope our work could inspire re-thinking the convention of dense prior in object detectors. The code is available at: https://github.com/PeizeSun/SparseR-CNN.

cs.CV↗

arXiv2021

A $k\cdot p$ effective Hamiltonian generator

$k\cdot p$ effective Hamiltonian is important for theoretical analysis in condensed matter physics. Based on the 'kdotp-symmetry' package, we develop an upgraded package named 'kdotp-generator'. This generator takes in arbitrary magnetic symmetries with their representations and returns symmetry-allowed $k\cdot p$ Hamiltonians. Using this package, we calculate $k\cdot p$ Hamiltonians for irreducible corepresentations in 1651 magnetic space groups up to the third order, and their linear coupling to external fields including the electromagnetic field and the strain tensor. We hope that the package will facilitate related research in the future.

cond-mat.mes-hall↗

arXiv2021

Hybrid Beamforming Optimization for DOA Estimation Based on the CRB Analysis

Direction-of-arrival (DOA) estimation is one of the most demanding tasks for the millimeter wave (mmWave) communication of massive multiple-input multiple-output (MIMO) systems with the hybrid beamforming (HBF) architecture. In this paper, we focus on the optimization of the HBF matrix for receiving pilots to enhance the DOA estimation performance. Motivated by the fact that many existing DOA estimation algorithms can achieve the Cramér-Rao bound (CRB), we formulate the HBF optimization problem aiming at minimizing the CRB with the prior knowledge of the rough DOA range. Then, to tackle the problem with intractable non-convex constraint introduced by the analog beamformers, we propose an efficient manifold optimization (MO) based algorithm. Simulation results demonstrate the significant improvement of the proposed CRB-MO algorithm over the conventional random HBF algorithm, and provide insights for the HBF design in the beam training stage for practical applications.

cs.IT↗

arXiv2021

Hybrid Interference Mitigation Using Analog Prewhitening

This paper proposes a novel scheme for mitigating strong interferences, which is applicable to various wireless scenarios, including full-duplex wireless communications and uncoordinated heterogenous networks. As strong interferences can saturate the receiver's analog-to-digital converters (ADC), they need to be mitigated both before and after the ADCs, i.e., via hybrid processing. The key idea of the proposed scheme, namely the Hybrid Interference Mitigation using Analog Prewhitening (HIMAP), is to insert an M-input M-output analog phase shifter network (PSN) between the receive antennas and the ADCs to spatially prewhiten the interferences, which requires no signal information but only an estimate of the covariance matrix. After interference mitigation by the PSN prewhitener, the preamble can be synchronized, the signal channel response can be estimated, and thus a minimum mean squared error (MMSE) beamformer can be applied in the digital domain to further mitigate the residual interferences. The simulation results verify that the HIMAP scheme can suppress interferences 80dB stronger than the signal by using off-the-shelf phase shifters (PS) of 6-bit resolution.

eess.SP↗

arXiv2021

Topological classification and diagnosis in magnetically ordered electronic materials

We show that compositions of time-reversal and spatial symmetries, also known as the magnetic-space-group symmetries, protect topological invariants as well as surface states that are distinct from those of all preceding topological states. We obtain, by explicit and exhaustive construction, the topological classification of electronic band insulators that are magnetically ordered for each one of the 1421 magnetic space groups in three dimensions. We have also computed the symmetry-based indicators for each nontrivial class, and, by doing so, establish the complete mapping from symmetry representations to topological invariants.

cond-mat.mes-hall↗

arXiv2021

Robust Output Regulation and Reinforcement Learning-based Output Tracking Design for Unknown Linear Discrete-Time Systems

In this paper, we investigate the optimal output tracking problem for linear discrete-time systems with unknown dynamics using reinforcement learning and robust output regulation theory. This output tracking problem only allows to utilize the outputs of the reference system and the controlled system, rather than their states, and differs from most existing tracking results that depend on the state of the system. The optimal tracking problem is formulated into a linear quadratic regulation problem by proposing a family of dynamic discrete-time controllers. Then, it is shown that solving the output tracking problem is equivalent to solving output regulation equations, whose solution, however, requires the knowledge of the complete and accurate system dynamics. To remove such a requirement, an off-policy reinforcement learning algorithm is proposed using only the measured output data along the trajectories of the system and the reference output. By introducing re-expression error and analyzing the rank condition of the parameterization matrix, we ensure the uniqueness of the proposed RL based optimal control via output feedback.

math.DS↗

arXiv2021

Electron ptychography achieves atomic-resolution limits set by lattice vibrations

Transmission electron microscopes use electrons with wavelengths of a few picometers, potentially capable of imaging individual atoms in solids at a resolution ultimately set by the intrinsic size of an atom. Unfortunately, due to imperfections in the imaging lenses and multiple scattering of electrons in the sample, the image resolution reached is 3 to 10 times worse. Here, by inversely solving the multiple scattering problem and overcoming the aberrations of the electron probe using electron ptychography to recover a linear phase response in thick samples, we demonstrate an instrumental blurring of under 20 picometers. The widths of atomic columns in the measured electrostatic potential are now no longer limited by the imaging system, but instead by the thermal fluctuations of the atoms. We also demonstrate that electron ptychography can potentially reach a sub-nanometer depth resolution and locate embedded atomic dopants in all three dimensions with only a single projection measurement.

cond-mat.mtrl-sci↗

arXiv2020

Optimising SARS-CoV-2 pooled testing strategies on social networks for low-resource settings

Controlling the COVID-19 pandemic is an urgent global challenge. The rapid geographic spread of SARS-CoV-2 directly reflects the social structure. Before effective vaccines and treatments are widely available, we have to rely on alternative, non-pharmaceutical interventions, including frequency testing, contact tracing, social distancing, mask wearing, and hand-washing, as public health practises to slow down the spread of the disease. However frequent testing is the key in the absence of any alternative. We propose a network approach to determine the optimal low resources setting oriented pool testing strategies that identifies infected individuals in a small number of tests and few rounds of testing, at low prevalence of the virus. We simulate stochastic infection curves on societies under quarantine. Allowing some social interaction is possible to keep the COVID-19 curve flat. However, similar results can be strategically obtained searching and isolating infected persons to preserve a healthier social structure. Here, we analyze which are the best strategies to contain the virus applying an algorithm that combine samples and testing them in groups [1]. A relevant parameter to keep infection curves flat using this algorithm is the dairy frequency of testing at zones where a high infection rate is reported. On the other hand, the algorithm efficiency is low for random search of infected people.

physics.soc-ph↗

arXiv2020

Transferring Orbital Angular Momentum to an Electron Beam Reveals Toroidal and Chiral Order

Orbital angular momentum and torque transfer play central roles in a wide range of magnetic textures and devices including skyrmions and spin-torque electronics(1-4). Analogous topological structures are now also being explored in ferroelectrics, including polarization vortex arrays in ferroelectric/dielectric superlattices(5). Unlike magnetic toroidal order, electric toroidal order does not couple directly to linear external fields. To develop a mechanism that can control switching in polarization vortices, we utilize a high-energy electron beam and show that transverse currents are generated by polar order in the ballistic limit. We find that the presence of an electric toroidal moment in a ferro-rotational phase transfers a measurable torque and orbital angular momentum to the electron beam. Furthermore, we find that the complex polarization patterns, observed in these heterostructures, are microscopically chiral with a non-trivial axial component of the polarization. This chirality opens the door for the coupling of ferroelectric and optical properties.

cond-mat.mtrl-sci↗

arXiv2020

A novel method to test particle ordering and final state alignment in helicity formalism

In this article, the importance is demonstrated of a proper choice of reference particles for decay angle definitions, when constructing partial-wave amplitude of multi-body decays using helicity formalism. This issue is often ignored in the standard use case of the helicity formalism. A new technique is proposed to determine the correct particle ordering, and can also be used as a generalized method to calculate the rotation operators used for the final-state alignment between different decay chains. Numerical validations are also performed to support the arguments and new technique proposed in this article.

hep-ex↗