arXiv subjects

Yong Zhao

Publications and source records attributed to Yong Zhao.

At least 181 records · Page 10Linked to original sources

arXiv2020

Reverse-engineering Bar Charts Using Neural Networks

Reverse-engineering bar charts extracts textual and numeric information from the visual representations of bar charts to support application scenarios that require the underlying information. In this paper, we propose a neural network-based method for reverse-engineering bar charts. We adopt a neural network-based object detection model to simultaneously localize and classify textual information. This approach improves the efficiency of textual information extraction. We design an encoder-decoder framework that integrates convolutional and recurrent neural networks to extract numeric information. We further introduce an attention mechanism into the framework to achieve high accuracy and robustness. Synthetic and real-world datasets are used to evaluate the effectiveness of the method. To the best of our knowledge, this work takes the lead in constructing a complete neural network-based method of reverse-engineering bar charts.

cs.CV↗

arXiv2020

Proton spin after 30 years: what we know and what we don't?

More than three decades has passed since the European Muon Collaboration published the first surprising result on the spin structure of the proton. Much theoretical and experimental progress has been made in understanding the origins of the proton spin. In this review, we will discuss what we have learned so far, what are still missing, and what we shall expect to learn from the upcoming experiments including JLab 12 GeV and Electron-Ion Collider. In particular, we focus on first principles calculations and experimental measurements of the total gluon helicity $ΔG$, and quark and gluon orbital angular momenta.

hep-ph↗

arXiv2020

Lattice Thermal Conductivity Prediction using Symbolic Regression and Machine Learning

Prediction models of lattice thermal conductivity have wide applications in the discovery of thermoelectrics, thermal barrier coatings, and thermal management of semiconductors. kL is notoriously difficult to predict. While classic models such as the Debye-Callaway model and the Slack model have been used to approximate the kL of inorganic compounds, their accuracy is far from being satisfactory. Herein, we propose a genetic programming based Symbolic Regression approach for explicit kL models and compare it with Multi-Layer Perceptron neural networks and a Random Forest Regressor using a hybrid cross-validation approach including both K-Fold CV and holdout validation. Four formulae have been discovered by our symbolic regression approach that outperform the Slack formula as evaluated on our dataset. Through the analysis of our models' performance and the formulae generated, we found that the trained formulae successfully reproduce the correct physical law that governs the lattice thermal conductivity of materials. We also identified that extrapolation prediction remains to be a key issue in both symbolic regression and regular machine learning methods and find the distribution of the samples place a key role in training a prediction model with high generalization capability.

cond-mat.mtrl-sci↗

arXiv2020

Collins-Soper Kernel for TMD Evolution from Lattice QCD

The Collins-Soper kernel relates transverse momentum-dependent parton distribution functions (TMDPDFs) at different energy scales. For small parton transverse momentum $q_T\sim Λ_\text{QCD}$, this kernel is non-perturbative and can only be determined with controlled uncertainties through experiment or first-principles calculations. This work presents the first exploratory determination of the Collins-Soper kernel using the lattice formulation of Quantum Chromodynamics. In a quenched calculation, the $N_f=0$ kernel is determined at scales in the range 250 MeV $< q_T < 2$ GeV, and an analysis of the remaining systematic uncertainties is undertaken.

hep-lat↗

arXiv2020

Cross-Channel Intragroup Sparsity Neural Network

Modern deep neural networks rely on overparameterization to achieve state-of-the-art generalization. But overparameterized models are computationally expensive. Network pruning is often employed to obtain less demanding models for deployment. Fine-grained pruning removes individual weights in parameter tensors and can achieve a high model compression ratio with little accuracy degradation. However, it introduces irregularity into the computing dataflow and often does not yield improved model inference efficiency in practice. Coarse-grained model pruning, while realizing satisfactory inference speedup through removal of network weights in groups, e.g. an entire filter, often lead to significant accuracy degradation. This work introduces the cross-channel intragroup (CCI) sparsity structure, which can prevent the inference inefficiency of fine-grained pruning while maintaining outstanding model performance. We then present a novel training algorithm designed to perform well under the constraint imposed by the CCI-Sparsity. Through a series of comparative experiments we show that our proposed CCI-Sparsity structure and the corresponding pruning algorithm outperform prior art in inference efficiency by a substantial margin given suited hardware acceleration in the future.

cs.LG↗

arXiv2020

Parton distribution function for the gluon condensate

Motivated by the desire to understand the nucleon mass structure in terms of light-cone distributions, we introduce the twist-four parton distribution function $F(x)$ whose first moment is the gluon condensate in the nucleon. We present the equation of motion relations for $F(x)$ and discuss the possible existence of the delta function (`zero mode') contribution at $x=0$. We also perform one-loop calculations for quark and gluon targets.

hep-ph↗

arXiv2020

Probing Nucleons and Nuclei in High Energy Collisions

This volume is a collection of contributions for the 7-week program "Probing Nucleons and Nuclei in High Energy Collisions" that was held at the Institute for Nuclear Theory in Seattle, WA, USA, from October 1 until November 16, 2018. The program was dedicated to the physics of the Electron Ion Collider (EIC), the world's first polarized electron-nucleon (ep) and electron-nucleus (eA) collider to be constructed in the USA. These proceedings are organized by chapters, corresponding to the weeks of the program: Week I, Generalized parton distributions; Week II, Transverse spin and TMDs; Week III, Longitudinal spin; Week IV, Symposium week; Weeks V & VI, eA collisions; Week VII, pA and AA collisions. We hope these proceedings will be useful to readers as a compilation of EIC-related science at the end of the second decade of the XXI century.

hep-ph↗

arXiv2020

One-loop Matching for Spin-Dependent Quasi-TMDs

Transverse momentum dependent parton distribution functions (TMDPDFs) provide a unique probe of the three-dimensional spin structure of hadrons. We construct spin-dependent quasi-TMDPDFs that are amenable to lattice QCD calculations and that can be used to determine spin-dependent TMDPDFs. We calculate the short-distance coefficients connecting spin-dependent TMDPDFs and quasi-TMDPDFs at one-loop order. We find that the helicity and transversity distributions have the same coefficient as the unpolarized TMDPDF. We also argue that the same is true for pretzelosity and that this spin universality of the matching will hold to all orders in $α_s$. Thus, it is possible to calculate ratios of these distributions as a function of longitudinal momentum and transverse position utilizing simpler Wilson line paths than have previously been considered.

hep-ph↗

arXiv2020

Machine Learning based prediction of noncentrosymmetric crystal materials

Noncentrosymmetric materials play a critical role in many important applications such as laser technology, communication systems,quantum computing, cybersecurity, and etc. However, the experimental discovery of new noncentrosymmetric materials is extremely difficult. Here we present a machine learning model that could predict whether the composition of a potential crystalline structure would be centrosymmetric or not. By evaluating a diverse set of composition features calculated using matminer featurizer package coupled with different machine learning algorithms, we find that Random Forest Classifiers give the best performance for noncentrosymmetric material prediction, reaching an accuracy of 84.8% when evaluated with 10 fold cross-validation on the dataset with 82,506 samples extracted from Materials Project. A random forest model trained with materials with only 3 elements gives even higher accuracy of 86.9%. We apply our ML model to screen potential noncentrosymmetric materials from 2,000,000 hypothetical materials generated by our inverse design engine and report the top 20 candidate noncentrosymmetric materials with 2 to 4 elements and top 20 borate candidates

physics.comp-ph↗

arXiv2020

Predicting Elastic Properties of Materials from Electronic Charge Density Using 3D Deep Convolutional Neural Networks

Materials representation plays a key role in machine learning based prediction of materials properties and new materials discovery. Currently both graph and 3D voxel representation methods are based on the heterogeneous elements of the crystal structures. Here, we propose to use electronic charge density (ECD) as a generic unified 3D descriptor for materials property prediction with the advantage of possessing close relation with the physical and chemical properties of materials. We developed an ECD based 3D convolutional neural networks (CNNs) for predicting elastic properties of materials, in which CNNs can learn effective hierarchical features with multiple convolving and pooling operations. Extensive benchmark experiments over 2,170 Fm-3m face-centered-cubic (FCC) materials show that our ECD based CNNs can achieve good performance for elasticity prediction. Especially, our CNN models based on the fusion of elemental Magpie features and ECD descriptors achieved the best 5-fold cross-validation performance. More importantly, we showed that our ECD based CNN models can achieve significantly better extrapolation performance when evaluated over non-redundant datasets where there are few neighbor training samples around test samples. As additional validation, we evaluated the predictive performance of our models on 329 materials of space group Fm-3m by comparing to DFT calculated values, which shows better prediction power of our model for bulk modulus than shear modulus. Due to the unified representation power of ECD, it is expected that our ECD based CNN approach can also be applied to predict other physical and chemical properties of crystalline materials.

cond-mat.mtrl-sci↗

arXiv2020

Nonperturbative renormalization of staple-shaped Wilson line operators in lattice QCD

Quark bilinear operators with staple-shaped Wilson lines are used to study transverse-momentum-dependent parton distribution functions (TMDPDFs) from lattice quantum chromodynamics (QCD). Here, the renormalization factors for the isovector operators, including all mixings between operators with different Dirac structures, are computed nonperturbatively in the regularization-independent momentum subtraction scheme for the first time. This study is undertaken in quenched QCD with three different lattice spacings. With Wilson flow applied to the gauge fields in the calculations, the operator mixing pattern due to chiral symmetry breaking with the lattice regularization is found to be significantly different from that predicted by one-loop lattice perturbation theory calculations. These results constitute a critical step towards the systematic extraction of TMDPDFs from lattice QCD.

hep-lat↗

arXiv2020

Pion valence quark PDF from lattice QCD

We present lattice results on the valence-quark structure of the pion using a coordinate space method within the framework of Large Momentum Effective Theory (LaMET). In this method one relies on the matrix elements of a Euclidean correlator in boosted hadronic states, which have an operator product expansion at short distance that allows us to extract the moments of PDFs. We renormalize the Euclidean correlator by forming the reduced Ioffe-time distribution (rITD), and reconstruct the second and fourth moments of the pion PDF by taking into account of QCD evolution effects.

hep-lat↗

arXiv2020

Dedge-AGMNet:an effective stereo matching network optimized by depth edge auxiliary task

To improve the performance in ill-posed regions, this paper proposes an atrous granular multi-scale network based on depth edge subnetwork(Dedge-AGMNet). According to a general fact, the depth edge is the binary semantic edge of instance-sensitive. This paper innovatively generates the depth edge ground-truth by mining the semantic and instance dataset simultaneously. To incorporate the depth edge cues efficiently, our network employs the hard parameter sharing mechanism for the stereo matching branch and depth edge branch. The network modifies SPP to Dedge-SPP, which fuses the depth edge features to the disparity estimation network. The granular convolution is extracted and extends to 3D architecture. Then we design the AGM module to build a more suitable structure. This module could capture the multi-scale receptive field with fewer parameters. Integrating the ranks of different stereo datasets, our network outperforms other stereo matching networks and advances state-of-the-art performances on the Sceneflow, KITTI 2012 and KITTI 2015 benchmark datasets.

cs.CV↗

arXiv2020

Renormalization and Matching for the Collins-Soper Kernel from Lattice QCD

The Collins-Soper kernel, which governs the energy evolution of transverse-momentum dependent parton distribution functions (TMDPDFs), is required to accurately predict Drell-Yan like processes at small transverse momentum, and is a key ingredient for extracting TMDPDFs from experiment. Earlier we proposed a method to calculate this kernel from ratios of the so-called quasi-TMDPDFs determined with lattice QCD, which are defined as hadronic matrix elements of staple-shaped Euclidean Wilson line operators. Here we provide the one-loop renormalization of these operators in a regularization-independent momentum subtraction (RI$^\prime$/MOM) scheme, as well as the conversion factor from the RI$^\prime$/MOM-renormalized quasi-TMDPDF to the $\overline{\rm MS}$ scheme. We also propose a procedure for calculating the Collins-Soper kernel directly from position space correlators, which simplifies the lattice determination.

hep-ph↗

arXiv2020

Global Attention based Graph Convolutional Neural Networks for Improved Materials Property Prediction

Machine learning (ML) methods have gained increasing popularity in exploring and developing new materials. More specifically, graph neural network (GNN) has been applied in predicting material properties. In this work, we develop a novel model, GATGNN, for predicting inorganic material properties based on graph neural networks composed of multiple graph-attention layers (GAT) and a global attention layer. Through the application of the GAT layers, our model can efficiently learn the complex bonds shared among the atoms within each atom's local neighborhood. Subsequently, the global attention layer provides the weight coefficients of each atom in the inorganic crystal material which are used to considerably improve our model's performance. Notably, with the development of our GATGNN model, we show that our method is able to both outperform the previous models' predictions and provide insight into the crystallization of the material.

physics.comp-ph↗

arXiv2020

Impact of JD Bernal Thoughts in the Science of Science upon China: Implications for Quantitative Studies of Science Today

John Desmond Bernal (1901-1970) was one of the most eminent scientists in molecular biology, and also regarded as the founding father of the Science of Science. His book The Social Function of Science laid the theoretical foundations for the discipline. In this article, we summarize four chief characteristics of his ideas in the Science of Science: the socio-historical perspective, theoretical models, qualitative and quantitative approaches, and studies of science planning and policy. China has constantly reformed its scientific and technological system based on research evidence of the Science of Science. Therefore, we analyze the impact of Bernal Science-of-Science thoughts on the development of Science of Science in China, and discuss how they might be usefully taken still further in quantitative studies of science.

physics.soc-ph↗

arXiv2019

Advances in Online Audio-Visual Meeting Transcription

This paper describes a system that generates speaker-annotated transcripts of meetings by using a microphone array and a 360-degree camera. The hallmark of the system is its ability to handle overlapped speech, which has been an unsolved problem in realistic settings for over a decade. We show that this problem can be addressed by using a continuous speech separation approach. In addition, we describe an online audio-visual speaker diarization method that leverages face tracking and identification, sound source localization, speaker identification, and, if available, prior speaker information for robustness to various real world challenges. All components are integrated in a meeting transcription framework called SRD, which stands for "separate, recognize, and diarize". Experimental results using recordings of natural meetings involving up to 11 attendees are reported. The continuous speech separation improves a word error rate (WER) by 16.1% compared with a highly tuned beamformer. When a complete list of meeting attendees is available, the discrepancy between WER and speaker-attributed WER is only 1.0%, indicating accurate word-to-speaker association. This increases marginally to 1.6% when 50% of the attendees are unknown to the system.

eess.AS↗

arXiv2019

Unpolarized isovector quark distribution function from Lattice QCD: A systematic analysis of renormalization and matching

We present a detailed Lattice QCD study of the unpolarized isovector quark Parton Distribution Function (PDF) using large-momentum effective theory framework. We choose a quasi-PDF defined by a spatial correlator which is free from mixing with other operators of the same dimension. In the lattice simulation, we use a Gaussian-momentum-smeared source at $M_π=356$ MeV and $P_z \in \{1.8,2.3\}$ GeV. To control the systematics associated with the excited states, we explore {five different source-sink separations}. The nonperturbative renormalization is conducted in a regularization-independent momentum subtraction scheme, and the matching between the renormalized quasi-PDF and $\bar{\rm MS}$ PDF is calculated based on perturbative QCD up to one-loop order. Systematic errors due to renormalization and perturbative matching are also analyzed in detail. Our results for lightcone PDF are in reasonable agreement with the latest phenomenological analysis.

hep-lat↗