arXiv subjects

Zheng Li

Publications and source records attributed to Zheng Li.

At least 379 records · Page 21Linked to original sources

arXiv2021

s-wave superconductivity in kagome metal CsV$_{3}$Sb$_{5}$ revealed by $^{121/123}$Sb NQR and $^{51}$V NMR measurements

We report $^{121/123}$Sb nuclear quadrupole resonance (NQR) and $^{51}$V nuclear magnetic resonance (NMR) measurements on kagome metal CsV$_3$Sb$_5$ with $T_{\rm c}=2.5$ K. Both $^{51}$V NMR spectra and $^{121/123}$Sb NQR spectra split after a charge density wave (CDW) transition, which demonstrates a commensurate CDW state. The coexistence of the high temperature phase and the CDW phase between $91$ K and $94$ K manifests that it is a first order phase transition. At low temperature, electric-field-gradient fluctuations diminish and magnetic fluctuations become dominant. Superconductivity emerges in the charge order state. Knight shift decreases and $1/T_{1}T$ shows a Hebel--Slichter coherence peak just below $T_{\rm c}$, indicating that CsV$_3$Sb$_5$ is an s-wave superconductor.

cond-mat.supr-con↗

arXiv2021

Two-Stage Self-Supervised Cycle-Consistency Network for Reconstruction of Thin-Slice MR Images

The thick-slice magnetic resonance (MR) images are often structurally blurred in coronal and sagittal views, which causes harm to diagnosis and image post-processing. Deep learning (DL) has shown great potential to re-construct the high-resolution (HR) thin-slice MR images from those low-resolution (LR) cases, which we refer to as the slice interpolation task in this work. However, since it is generally difficult to sample abundant paired LR-HR MR images, the classical fully supervised DL-based models cannot be effectively trained to get robust performance. To this end, we propose a novel Two-stage Self-supervised Cycle-consistency Network (TSCNet) for MR slice interpolation, in which a two-stage self-supervised learning (SSL) strategy is developed for unsupervised DL network training. The paired LR-HR images are synthesized along the sagittal and coronal directions of input LR images for network pretraining in the first-stage SSL, and then a cyclic in-terpolation procedure based on triplet axial slices is designed in the second-stage SSL for further refinement. More training samples with rich contexts along all directions are exploited as guidance to guarantee the improved in-terpolation performance. Moreover, a new cycle-consistency constraint is proposed to supervise this cyclic procedure, which encourages the network to reconstruct more realistic HR images. The experimental results on a real MRI dataset indicate that TSCNet achieves superior performance over the conventional and other SSL-based algorithms, and obtains competitive quali-tative and quantitative results compared with the fully supervised algorithm.

eess.IV↗

arXiv2021

Phoneme-aware and Channel-wise Attentive Learning for Text DependentSpeaker Verification

This paper proposes a multi-task learning network with phoneme-aware and channel-wise attentive learning strategies for text-dependent Speaker Verification (SV). In the proposed structure, the frame-level multi-task learning along with the segment-level adversarial learning is adopted for speaker embedding extraction. The phoneme-aware attentive pooling is exploited on frame-level features in the main network for speaker classifier, with the corresponding posterior probability for the phoneme distribution in the auxiliary subnet. Further, the introduction of Squeeze and Excitation (SE-block) performs dynamic channel-wise feature recalibration, which improves the representational ability. The proposed method exploits speaker idiosyncrasies associated with pass-phrases, and is further improved by the phoneme-aware attentive pooling and SE-block from temporal and channel-wise aspects, respectively. The experiments conducted on RSR2015 Part 1 database confirm that the proposed system achieves outstanding results for textdependent SV.

cs.SD↗

arXiv2021

Additive Phoneme-aware Margin Softmax Loss for Language Recognition

This paper proposes an additive phoneme-aware margin softmax (APM-Softmax) loss to train the multi-task learning network with phonetic information for language recognition. In additive margin softmax (AM-Softmax) loss, the margin is set as a constant during the entire training for all training samples, and that is a suboptimal method since the recognition difficulty varies in training samples. In additive angular margin softmax (AAM-Softmax) loss, the additional angular margin is set as a costant as well. In this paper, we propose an APM-Softmax loss for language recognition with phoneitc multi-task learning, in which the additive phoneme-aware margin is automatically tuned for different training samples. More specifically, the margin of language recognition is adjusted according to the results of phoneme recognition. Experiments are reported on Oriental Language Recognition (OLR) datasets, and the proposed method improves AM-Softmax loss and AAM-Softmax loss in different language recognition testing conditions.

cs.SD↗

arXiv2021

Unsupervised Cross-lingual Adaptation for Sequence Tagging and Beyond

Cross-lingual adaptation with multilingual pre-trained language models (mPTLMs) mainly consists of two lines of works: zero-shot approach and translation-based approach, which have been studied extensively on the sequence-level tasks. We further verify the efficacy of these cross-lingual adaptation approaches by evaluating their performances on more fine-grained sequence tagging tasks. After re-examining their strengths and drawbacks, we propose a novel framework to consolidate the zero-shot approach and the translation-based approach for better adaptation performance. Instead of simply augmenting the source data with the machine-translated data, we tailor-make a warm-up mechanism to quickly update the mPTLMs with the gradients estimated on a few translated data. Then, the adaptation approach is applied to the refined parameters and the cross-lingual transfer is performed in a warm-start way. The experimental results on nine target languages demonstrate that our method is beneficial to the cross-lingual adaptation of various sequence tagging tasks.

cs.CL↗

arXiv2021

On Linearly Precoded Rate Splitting for Gaussian MIMO Broadcast Channels

In this paper, we consider a general K-user Gaussian multiple-input multiple-output (MIMO) broadcast channel (BC). We assume that the channel state is deterministic and known to all the nodes. While the private-message capacity region is well known to be achievable with dirty paper coding (DPC), we are interested in the simpler linearly precoded transmission schemes. In particular, we focus on linear precoding schemes combined with rate-splitting (RS). First, we derive an achievable rate region with minimum mean square error (MMSE) precoding at the transmitter and joint decoding of the sub-messages at the receivers. Then, we study the achievable sum rate of this scheme and obtain two findings: 1) an analytically tractable upper bound on the sum rate that is shown numerically to be a close approximation, and 2) how to reduce the number of active streams -- crucial to the overall complexity -- while preserving the sum rate to within a constant loss. The latter results in two practical algorithms: a stream elimination algorithm and a stream ordering algorithm. Finally, we investigate the constant-gap optimality of linearly precoded RS with respect to the capacity. Our result reveals that, while the achievable rate of linear precoding alone can be arbitrarily far from the capacity, the introduction of RS can help achieve the capacity region to within a constant gap in the two-user case. Nevertheless, we prove that the RS scheme's constant-gap optimality does not extend to the three-user case. Specifically, we show, through a pathological example, that the gap between the sum rate and the sum capacity can be unbounded.

cs.IT↗

arXiv2021

SUOD: Accelerating Large-Scale Unsupervised Heterogeneous Outlier Detection

Outlier detection (OD) is a key machine learning (ML) task for identifying abnormal objects from general samples with numerous high-stake applications including fraud detection and intrusion detection. Due to the lack of ground truth labels, practitioners often have to build a large number of unsupervised, heterogeneous models (i.e., different algorithms with varying hyperparameters) for further combination and analysis, rather than relying on a single model. How to accelerate the training and scoring on new-coming samples by outlyingness (referred as prediction throughout the paper) with a large number of unsupervised, heterogeneous OD models? In this study, we propose a modular acceleration system, called SUOD, to address it. The proposed system focuses on three complementary acceleration aspects (data reduction for high-dimensional data, approximation for costly models, and taskload imbalance optimization for distributed environment), while maintaining performance accuracy. Extensive experiments on more than 20 benchmark datasets demonstrate SUOD's effectiveness in heterogeneous OD acceleration, along with a real-world deployment case on fraudulent claim analysis at IQVIA, a leading healthcare firm. We open-source SUOD for reproducibility and accessibility.

cs.LG↗

arXiv2021

Long Live The Image: Container-Native Data Persistence in Production

Containerization plays a crucial role in the de facto technology stack for implementing microservices architecture (each microservice has its own database in most cases). Nevertheless, there are still fierce debates on containerizing production databases, mainly due to the data persistence issues and concerns. Driven by a project of refactoring an Automated Machine Learning system, this research proposes the container-native data persistence as a conditional solution to running database containers in production. In essence, the proposed solution distinguishes the stateless data access (i.e. reading) from the stateful data processing (i.e. creating, updating, and deleting) in databases. A master database handles the stateful data processing and dumps database copies for building container images, while the database containers will keep stateless at runtime, based on the preloaded dump in the image. Although there are delays in the state/image update propagation, this solution is particularly suitable for the read-only, the eventual consistency, and the asynchronous processing scenarios. Moreover, with optimal tuning (e.g., disabling locking), the portability and performance gains of a read-only database container would outweigh the performance loss in accessing data across the underlying image layers.

cs.DB↗

arXiv2021

Stop Building Castles on a Swamp! The Crisis of Reproducing Automatic Search in Evidence-based Software Engineering

The evidence-based approach has increasingly been employed to synthesize empirical findings from the primary research in software engineering. Nevertheless, the reproducibility of evidence-based software engineering (EBSE) studies seems to be underemphasized. In our investigation into the automatic search of 311 sample studies, more than 50% of the search strings are not reusable; about 87.5% of the search activities (e.g., search field settings) are unrepeatable; and more than 95% of the whole automatic search implementations are unreproducible. Considering that searching is a cornerstone of an EBSE study, we are afraid that the reproducibility of the current secondary research could be worse than we can imagine. By analyzing and reporting the root causes of the aforementioned observations, we urge collaboration and cooperation among all the stakeholders in our community to improve the research reproducibility in EBSE.

cs.SE↗

arXiv2021

On a Factorial Knowledge Architecture for Data Science-powered Software Engineering

Given the data-intensive and collaborative trend in science, the software engineering community also pays increasing attention to obtaining valuable and useful insights from data repositories. Nevertheless, applying data science to software engineering (e.g., mining software repositories) can be blindfold and meaningless, if lacking a suitable knowledge architecture (KA). By observing that software engineering practices are generally recorded through a set of factors (e.g., programmer capacity, different environmental conditions, etc.) involved in various software project aspects, we propose a factor-based hierarchical KA of software engineering to help maximize the value of software repositories and inspire future software data-driven studies. In particular, it is the organized factors and their relationships that help guide software engineering knowledge mining, while the mined knowledge will in turn be indexed/managed through the relevant factors and their interactions. This paper explains our idea about the factorial KA and concisely demonstrates a KA component, i.e. the early-version KA of software product engineering. Once fully scoped, this proposed KA will supplement the well-known SWEBOK in terms of both the factor-centric knowledge management and the coverage/implication of potential software engineering knowledge.

cs.SE↗

arXiv2021

Design and Control of a Highly Redundant Rigid-Flexible Coupling Robot to Assist the COVID-19 Oropharyngeal-Swab Sampling

The outbreak of novel coronavirus pneumonia (COVID-19) has caused mortality and morbidity worldwide. Oropharyngeal-swab (OP-swab) sampling is widely used for the diagnosis of COVID-19 in the world. To avoid the clinical staff from being affected by the virus, we developed a 9-degree-of-freedom (DOF) rigid-flexible coupling (RFC) robot to assist the COVID-19 OP-swab sampling. This robot is composed of a visual system, UR5 robot arm, micro-pneumatic actuator and force-sensing system. The robot is expected to reduce risk and free up the clinical staff from the long-term repetitive sampling work. Compared with a rigid sampling robot, the developed force-sensing RFC robot can facilitate OP-swab sampling procedures in a safer and softer way. In addition, a varying-parameter zeroing neural network-based optimization method is also proposed for motion planning of the 9-DOF redundant manipulator. The developed robot system is validated by OP-swab sampling on both oral cavity phantoms and volunteers.

cs.RO↗

arXiv2021

Superresolving second-order correlation imaging using synthesized colored noise speckles

We present a novel method to synthesize non-trivial speckles that can enable superresolving second-order correlation imaging. The speckles acquire a unique anti-correlation in the spatial intensity fluctuation by introducing the blue noise spectrum to the input light fields through amplitude modulation. Illuminating objects with the blue noise speckle patterns can lead to a sub-diffraction limit imaging system with a resolution more than three times higher than first-order imaging, which is comparable to the resolving power of ninth order correlation imaging with thermal light. Our method opens a new route towards non-trivial speckle generation by tailoring amplitudes of the input light fields and provides a versatile scheme for constructing superresolving imaging and microscopy systems without invoking complicated higher-order correlations.

physics.optics↗

arXiv2021

A one-dimensional flow model enhanced by machine learning for simulation of vocal fold vibration

We describe a one-dimensional (1D) unsteady and viscous flow model that is derived from the momentum and mass conservation equations, and to enhance this physics-based model, we use a machine learning approach to determine the unknown modeling parameters. Specifically, we first construct an idealized larynx model and perform ten cases of three-dimensional (3D) fluid--structure interaction (FSI) simulations. The flow data are then extracted to train the 1D flow model using a sparse identification approach for nonlinear dynamical systems. As a result of training, we obtain the analytical expressions for the entrance effect and pressure loss in the glottis, which are then incorporated in the flow model to conveniently handle different glottal shapes due to vocal fold vibration. We apply the enhanced 1D flow model in the FSI simulation of both idealized vocal fold geometries and subject-specific anatomical geometries reconstructed from the MRI images of rabbits' larynges. The 1D flow model is evaluated in both of these setups and is shown to have robust performance. Therefore, it provides a fast simulation tool superior to the previous 1D models.

physics.flu-dyn↗

arXiv2021

Reconstruction of Quantitative Susceptibility Maps from Phase of Susceptibility Weighted Imaging with Cross-Connected $Ψ$-Net

Quantitative Susceptibility Mapping (QSM) is a new phase-based technique for quantifying magnetic susceptibility. The existing QSM reconstruction methods generally require complicated pre-processing on high-quality phase data. In this work, we propose to explore a new value of the high-pass filtered phase data generated in susceptibility weighted imaging (SWI), and develop an end-to-end Cross-connected $Ψ$-Net (C$Ψ$-Net) to reconstruct QSM directly from these phase data in SWI without additional pre-processing. C$Ψ$-Net adds an intermediate branch in the classical U-Net to form a $Ψ$-like structure. The specially designed dilated interaction block is embedded in each level of this branch to enlarge the receptive fields for capturing more susceptibility information from a wider spatial range of phase images. Moreover, the crossed connections are utilized between branches to implement a multi-resolution feature fusion scheme, which helps C$Ψ$-Net capture rich contextual information for accurate reconstruction. The experimental results on a human dataset show that C$Ψ$-Net achieves superior performance in our task over other QSM reconstruction algorithms.

eess.IV↗

arXiv2021

Exploring Text-transformers in AAAI 2021 Shared Task: COVID-19 Fake News Detection in English

In this paper, we describe our system for the AAAI 2021 shared task of COVID-19 Fake News Detection in English, where we achieved the 3rd position with the weighted F1 score of 0.9859 on the test set. Specifically, we proposed an ensemble method of different pre-trained language models such as BERT, Roberta, Ernie, etc. with various training strategies including warm-up,learning rate schedule and k-fold cross-validation. We also conduct an extensive analysis of the samples that are not correctly classified. The code is available at:https://github.com/archersama/3rd-solution-COVID19-Fake-News-Detection-in-English.

cs.CL↗

arXiv2020

Tomographic imaging of complete quantum state of matter by ultrafast diffraction

With the ability to directly obtain the Wigner function and density matrix of photon states, quantum tomography (QT) has had a significant impact on quantum optics, quantum computing and quantum information. By an appropriate sequence of measurements on the evolution of each degreeof freedom (DOF), the full quantum state of the observed photonic system can be determined. The first proposal to extend the application of QT to reconstruction of complete quantum states of matter wavepackets had generated enormous interest in ultrafast diffraction imaging and pump-probe spectroscopy of molecules. This interest was elevated with the advent of ultrafast electron and X-ray diffraction techniques using electron accelerators and X-ray free electron lasers to add temporal resolution to the observed nuclear and electron distributions. However, the great interest in this area has been tempered by the illustration of an impossibility theorem, known as the dimension problem. Not being able to associate unitary evolution to every DOF of molecular motion, quantum tomography could not be used beyond 1D and categorically excludes most vibrational and all rotational motion of molecules. Here we present a theoretical advance to overcome the notorious dimension problem. Solving this challenging problem is important to push imaging molecular dynamics to the quantum limit. The new theory has solved this problem, which makes quantum tomography a truly useful methodology in ultrafast physics and enables the making of quantum version of a molecular movie. With the new theory, quantum tomography can be finally advanced to a sufficient level to become a general method for reconstructing quantum states of matter, without being limited in one dimension. Our new concept is demonstrated using a simulated dataset of ultrafast diffraction experiment of laser-aligned nitrogen molecules.

quant-ph↗

arXiv2020

Site-specific Interrogation of an Ionic Chiral Fragment During Photolysis Using an X-ray Free-Electron Laser

Short-wavelength free-electron lasers with their ultrashort pulses at high intensities have originated new approaches for tracking molecular dynamics from the vista of specific sites. X-ray pump X-ray probe schemes even allow to address individual atomic constituents with a 'trigger'-event that preludes the subsequent molecular dynamics while being able to selectively probe the evolving structure with a time-delayed second X-ray pulse. Here, we use a linearly polarized X-ray photon to trigger the photolysis of a prototypical chiral molecule, namely trifluoromethyloxirane (C$_3$H$_3$F$_3$O), at the fluorine K-edge at around 700 eV. The evolving fluorine-containing fragments are then probed by a second, circularly polarized X-ray pulse of higher photon energy in order to investigate the chemically shifted inner-shell electrons of the ionic motherfragment for their stereochemical sensitivity. We experimentally demonstrate and theoretically support how two-color X-ray pump X-ray probe experiments with polarization control enable XFELs as tools for chiral recognition.

physics.chem-ph↗

arXiv2020

Online Knowledge Distillation via Multi-branch Diversity Enhancement

Knowledge distillation is an effective method to transfer the knowledge from the cumbersome teacher model to the lightweight student model. Online knowledge distillation uses the ensembled prediction results of multiple student models as soft targets to train each student model. However, the homogenization problem will lead to difficulty in further improving model performance. In this work, we propose a new distillation method to enhance the diversity among multiple student models. We introduce Feature Fusion Module (FFM), which improves the performance of the attention mechanism in the network by integrating rich semantic information contained in the last block of multiple student models. Furthermore, we use the Classifier Diversification(CD) loss function to strengthen the differences between the student models and deliver a better ensemble result. Extensive experiments proved that our method significantly enhances the diversity among student models and brings better distillation performance. We evaluate our method on three image classification datasets: CIFAR-10/100 and CINIC-10. The results show that our method achieves state-of-the-art performance on these datasets.

cs.CV↗