Searcharxiv⌕ Search

arXiv subjects

Xiaoyu Wang

Publications and source records attributed to Xiaoyu Wang.

At least 163 records · Page 9Linked to original sources
arXiv2023

Influence of N,N,N-trimethyl-1-adamantyl ammonium (TMAda+) Structure Directing Agent on Al Pair Distributions and Features in Chabazite Zeolite

While organic structure directing agents (OSDAs) are well known to have a directional influence on the topology of a crystallizing zeolite, the relationship between OSDA charge and siting of aliovalent ions on a primarily siliceous framework is unclear. Here, we explore the relationship between OSDA orientation, Al3+ siting, and lattice energy, taking as a model system CHA zeolite occluded with N,N,N-trimethyl-1-adamantyl ammonium (TMAda+) at an Si/Al ratio of 11/1. We use density functional theory calculations to parametrize a fixed-charge classical model describing van der Waals and electrostatic interactions between framework and OSDA. We enumerate and explore all possible combinations of OSDA orientation and Al location (attending to Lowenstein's rule) within a 36 T-site supercell. We find that interaction energies vary over 60 kJ/double-six-ring-unit (d6r). Further, analysis of configurations reveals that energies are sensitive to Al-Al proximity, such that low energies are associated with Al3+ pairs in 8-membered rings and higher energies associated with Al3+ pairs in smaller 6- and 4-membered rings. Comparisons with Al siting inferred from CHA zeolite crystallized with TMAda+ suggests that these computed interaction energies are useful reporters of observed Al siting in CHA synthesized with TMAda+.

cond-mat.mtrl-sci↗
arXiv2023

Revisiting Bloch electrons in magnetic field: Hofstadter physics via hybrid Wannier states

We revisit the Hofstadter butterfly for a subset of topologically trivial Bloch bands arising from a continuum free electron Hamiltonian in a periodic lattice potential. We employ the recently developed procedure -- which was previously used to analyze the case of topologically non-trivial bands [\href{https://journals.aps.org/prb/abstract/10.1103/PhysRevB.106.L121111}{Phys. Rev. B \textbf{106}, L121111 (2022)}] -- to construct the finite field Hilbert space from the zero-field hybrid Wannier basis states. Such states are Bloch extended along one direction and exponentially localized along the other. The method is illustrated for square and triangular lattice potentials and is shown to reproduce all the main features of the Hofstadter spectrum obtained from a numerically exact Landau level expansion method. In the regime when magnetic length is much longer than the spatial extent of the hybrid Wannier state in the localized direction we recover the well known Harper equation. Because the method applies to both topologically trivial and non-trivial bands, it provides an alternative and efficient approach to moiré materials in magnetic field.

cond-mat.mes-hall↗
arXiv2023

The mass spectrum and strong decay properties of the charmed-strange mesons within Godfrey-Isgur model considering the coupled-channel effects

Motivated by the recently observed $D_{s0}(2590)$ state by LHCb, we investigate the mass spectrum and the strong decay properties of the charmed-strange mesons within Godfrey-Isgur model considering the coupled-channel effects. Our results support that $D_{s0}^*(2317)$ and $D_{s1}(2460)$ can be interpreted as the $D_{s}(1^3P_0)$ and $D_{s}(1^3P_1)$ states with larger $DK$ and $D^*K$ components, respectively, and $D_{s1}(2700)$, $D_{s1}(2536)$, $D^*_{s2}(2573)$, $D_{s1}^*(2860)$, $D_{s3}^*(2860)$, and $D_{sJ}^*(3040)$ can be well interpreted as the $D_s(2^3S_1)$, $D_s(1^1P_1)$, $D_s(1^3P_2)$, $D_s(1^3D_1)$, $D_s(1^3D_3)$, and $D_s(2^1P_1)$ states, respectively. Although, $D_{s0}(2590)$ mass is about 50 MeV less than our prediction for the $D_{s}(2^1S_0)$ state, its width is still in good agreement with the one of $D_{s}(2^1S_0)$. Therefore, $D_{s0}(2590)$ state needs to be further confirmed by the experimental measurements, and the more precise information about $D_{s0}(2590)$ will shed light on its assignment of $D_{s}(2^1S_0)$. Furthermore, we predict the masses and the strong decay properties of the charmed-strange mesons with masses around 3 GeV, which would be helpful to experimentally search for these states.

hep-ph↗
arXiv2023

Fast-ParC: Capturing Position Aware Global Feature for ConvNets and ViTs

Transformer models have made tremendous progress in various fields in recent years. In the field of computer vision, vision transformers (ViTs) also become strong alternatives to convolutional neural networks (ConvNets), yet they have not been able to replace ConvNets since both have their own merits. For instance, ViTs are good at extracting global features with attention mechanisms while ConvNets are more efficient in modeling local relationships due to their strong inductive bias. A natural idea that arises is to combine the strengths of both ConvNets and ViTs to design new structures. In this paper, we propose a new basic neural network operator named position-aware circular convolution (ParC) and its accelerated version Fast-ParC. The ParC operator can capture global features by using a global kernel and circular convolution while keeping location sensitiveness by employing position embeddings. Our Fast-ParC further reduces the O(n2) time complexity of ParC to O(n log n) using Fast Fourier Transform. This acceleration makes it possible to use global convolution in the early stages of models with large feature maps, yet still maintains the overall computational cost comparable with using 3x3 or 7x7 kernels. The proposed operation can be used in a plug-and-play manner to 1) convert ViTs to pure-ConvNet architecture to enjoy wider hardware support and achieve higher inference speed; 2) replacing traditional convolutions in the deep stage of ConvNets to improve accuracy by enlarging the effective receptive field. Experiment results show that our ParC op can effectively enlarge the receptive field of traditional ConvNets, and adopting the proposed op benefits both ViTs and ConvNet models on all three popular vision tasks, image classification, object

cs.CV↗
arXiv2023

Generalized Polyak Step Size for First Order Optimization with Momentum

In machine learning applications, it is well known that carefully designed learning rate (step size) schedules can significantly improve the convergence of commonly used first-order optimization algorithms. Therefore how to set step size adaptively becomes an important research question. A popular and effective method is the Polyak step size, which sets step size adaptively for gradient descent or stochastic gradient descent without the need to estimate the smoothness parameter of the objective function. However, there has not been a principled way to generalize the Polyak step size for algorithms with momentum accelerations. This paper presents a general framework to set the learning rate adaptively for first-order optimization methods with momentum, motivated by the derivation of Polyak step size. It is shown that the resulting techniques are much less sensitive to the choice of momentum parameter and may avoid the oscillation of the heavy-ball method on ill-conditioned problems. These adaptive step sizes are further extended to the stochastic settings, which are attractive choices for stochastic gradient descent with momentum. Our methods are demonstrated to be more effective for stochastic gradient methods than prior adaptive step size algorithms in large-scale machine learning tasks.

math.OC↗
arXiv2023

NAR-Former V2: Rethinking Transformer for Universal Neural Network Representation Learning

As more deep learning models are being applied in real-world applications, there is a growing need for modeling and learning the representations of neural networks themselves. An efficient representation can be used to predict target attributes of networks without the need for actual training and deployment procedures, facilitating efficient network deployment and design. Recently, inspired by the success of Transformer, some Transformer-based representation learning frameworks have been proposed and achieved promising performance in handling cell-structured models. However, graph neural network (GNN) based approaches still dominate the field of learning representation for the entire network. In this paper, we revisit Transformer and compare it with GNN to analyse their different architecture characteristics. We then propose a modified Transformer-based universal neural network representation learning model NAR-Former V2. It can learn efficient representations from both cell-structured networks and entire networks. Specifically, we first take the network as a graph and design a straightforward tokenizer to encode the network into a sequence. Then, we incorporate the inductive representation learning capability of GNN into Transformer, enabling Transformer to generalize better when encountering unseen architecture. Additionally, we introduce a series of simple yet effective modifications to enhance the ability of the Transformer in learning representation from graph structures. Our proposed method surpasses the GNN-based method NNLP by a significant margin in latency estimation on the NNLQP dataset. Furthermore, regarding accuracy prediction on the NASBench101 and NASBench201 datasets, our method achieves highly comparable performance to other state-of-the-art methods.

cs.LG↗
arXiv2023

Universal Object Detection with Large Vision Model

Over the past few years, there has been growing interest in developing a broad, universal, and general-purpose computer vision system. Such systems have the potential to address a wide range of vision tasks simultaneously, without being limited to specific problems or data domains. This universality is crucial for practical, real-world computer vision applications. In this study, our focus is on a specific challenge: the large-scale, multi-domain universal object detection problem, which contributes to the broader goal of achieving a universal vision system. This problem presents several intricate challenges, including cross-dataset category label duplication, label conflicts, and the necessity to handle hierarchical taxonomies. To address these challenges, we introduce our approach to label handling, hierarchy-aware loss design, and resource-efficient model training utilizing a pre-trained large vision model. Our method has demonstrated remarkable performance, securing a prestigious second-place ranking in the object detection track of the Robust Vision Challenge 2022 (RVC 2022) on a million-scale cross-dataset object detection benchmark. We believe that our comprehensive study will serve as a valuable reference and offer an alternative approach for addressing similar challenges within the computer vision community. The source code for our work is openly available at https://github.com/linfeng93/Large-UniDet.

cs.CV↗
arXiv2023

Quantum anomalous Hall state in a fluorinated 1T-MoSe$_2$ monolayer

The quantum anomalous Hall state with a large band gap and a high Chern number is significant for practical applications in spintronics. By performing first-principles calculations, we investigate electronic properties of the fully fluorinated 1T-MoSe$_{2}$ monolayer. Without considering the spin-orbit coupling, the band structure demonstrates single-spin semi-metallic properties and the trigonal warping around $K_{\pm}$ valleys. The introduction of the spin-orbit coupling opens considerable band gaps of $117.2$ meV around the two valleys, leading to a nontrivial quantum anomalous Hall state with a Chern number of $|C|=2$, which provides two chiral dissipationless transport channels from topological edge states and associated quantized anomalous Hall conductivity. In addition, an effective model is constructed to describe the low-energy physics of the monolayer. Our findings in the MoSe$_{2}$F$_{2}$ monolayer sheds light on large-gap quantum anomalous Hall states in two-dimensional materials with the chemical functionalization, and provides opportunities in designing low-power and noise-tolerant spintronic devices.

cond-mat.mes-hall↗
arXiv2023

Tunable Magnetism and Valleys in VSiZ$_3$ monolayers

Two-dimensional magnetism and valleys have recently emerged as two significant research areas, with intriguing properties and practical uses in advanced information technology. Considering the importance of these two areas and their couplings, controllable creations of both the magnetism and valley polarization are highly sought after. Based on first-principles calculations, we propose a new class of two-dimensional monolayers with a chemical formula of MAZ$_3$, which is viewed as a 2H-MZ$_2$ trilayer passivated by the A-Z bilayer on its one side. Taking VSiN$_3$ as an example, the MAZ$_3$ monolayers are found to exhibit tunable magnetism and valleys. For the intrinsic VSiN$_3$ monolayer, it is a non-magnetic semiconductor, with multiple degenerate valleys and trigonal warping near $K_\pm$ points in the band structure. Besides, the bands have spin splittings owing to the spin-orbit coupling. Under a moderate carrier doping, the monolayer becomes a Stoner ferromagnet, which enhances the spin splittings of the valence band and generates valley splittings. Moreover, the Berry curvature is valley contrasting, leading to distinct valley-spin related anomalous Hall currents as the doping concentration increases. Our work opens up new way to modulate the spin splittings and valley splittings via electric means, and provides opportunities for exploring advanced spintronic and valleytronic devices.

cond-mat.mtrl-sci↗
arXiv2023

Magnetically and electrically controllable valley splittings in MXene monolayers

The modulation of the valley structure in two-dimensional valley materials is vital in the field of valleytronics. The multiferroicity provides possibility for multiple modulations of the valley, including the magnetic and electric means. Based on the first-principle calculations, we study the valley properties and associated manipulations of multiferroic Co$_2$CF$_2$ monolayers with different stacking patterns. Our calculations show that the Co$_2$CF$_2$ monolayer in the H$^{\prime}$ phase is a ferrovalley material, with sizable valley splittings. By rotating the magnetization direction, the valley splittings can be tuned for both the magnitude and sign. The electric field, driving the reversal of the electric polarization, can also change the magnitude of the valley splittings. Besides, a metastable T$^{\prime}$ phase exhibits valley splittings as well, of which the magnitude and sign can be simultaneously controlled by applied magnetic and electric fields. These findings offer a practical way for realizing highly tunable valleys by multiferroic couplings.

cond-mat.mtrl-sci↗
arXiv2023

Incorporating increased variability in testing for cancer DNA methylation

Cancer development is associated with aberrant DNA methylation, including increased stochastic variability. Statistical tests for discovering cancer methylation biomarkers have focused on changes in mean methylation. To improve the power of detection, we propose to incorporate increased variability in testing for cancer differential methylation by two joint constrained tests: one for differential mean and increased variance, the other for increased mean and increased variance. To improve small sample properties, likelihood ratio statistics are developed, accounting for the variability in estimating the sample medians in the Levene test. Efficient algorithms were developed and implemented in DMVC function of R package DMtest. The proposed joint constrained tests were compared to standard tests and partial area under the curve (pAUC) for the receiver operating characteristic curve (ROC) in simulated datasets under diverse models. Application to the high-throughput methylome data in The Cancer Genome Atlas (TCGA) shows substantially increased yield of candidate CpG markers.

stat.ME↗
arXiv2023

DMseg: a Python algorithm for de novo detection of differentially or variably methylated regions

Detecting and assessing statistical significance of differentially methylated regions (DMRs) is a fundamental task in methylome association studies. While the average differential methylation in different phenotype groups has been the inferential focus, methylation changes in chromosomal regions may also present as differential variability, i.e., variably methylated regions (VMRs). Testing statistical significance of regional differential methylation is a challenging problem, and existing algorithms do not provide accurate type I error control for genome-wide DMR or VMR analysis. No algorithm has been publicly available for detecting VMRs. We propose DMseg, a Python algorithm with efficient DMR/VMR detection and significance assessment for array-based methylome data, and compare its performance to Bumphunter, a popular existing algorithm. Operationally, DMseg searches for DMRs or VMRs within CpG clusters that are adaptively determined by both gap distance and correlation between contiguous CpG sites in a microarray. Levene test was implemented for assessing differential variability of individual CpGs. A likelihood ratio statistic is proposed to test for a constant difference within CpGs in a DMR or VMR to summarize the evidence of regional difference. Using a stratified permutation scheme and pooling null distributions of LRTs from clusters with similar numbers of CpGs, DMseg provides accurate control of the type I error rate. In simulation experiments, DMseg shows superior power than Bumphunter to detect DMRs. Application to methylome data of Barrett's esophagus and esophageal adenocarcinoma reveals a number of DMRs and VMRs of biological interest.

stat.ME↗
arXiv2023

Perimeter Control Using Deep Reinforcement Learning: A Model-free Approach towards Homogeneous Flow Rate Optimization

Perimeter control maintains high traffic efficiency within protected regions by controlling transfer flows among regions to ensure that their traffic densities are below critical values. Existing approaches can be categorized as either model-based or model-free, depending on whether they rely on network transmission models (NTMs) and macroscopic fundamental diagrams (MFDs). Although model-based approaches are more data efficient and have performance guarantees, they are inherently prone to model bias and inaccuracy. For example, NTMs often become imprecise for a large number of protected regions, and MFDs can exhibit scatter and hysteresis that are not captured in existing model-based works. Moreover, no existing studies have employed reinforcement learning for homogeneous flow rate optimization in microscopic simulation, where spatial characteristics, vehicle-level information, and metering realizations -- often overlooked in macroscopic simulations -- are taken into account. To circumvent issues of model-based approaches and macroscopic simulation, we propose a model-free deep reinforcement learning approach that optimizes the flow rate homogeneously at the perimeter at the microscopic level. Results demonstrate that our model-free reinforcement learning approach without any knowledge of NTMs or MFDs can compete and match the performance of a model-based approach, and exhibits enhanced generalizability and scalability.

cs.LG↗
arXiv2023

Path Planning for Air-Ground Robot Considering Modal Switching Point Optimization

An innovative sort of mobility platform that can both drive and fly is the air-ground robot. The need for an agile flight cannot be satisfied by traditional path planning techniques for air-ground robots. Prior studies had mostly focused on improving the energy efficiency of paths, seldom taking the seeking speed and optimizing take-off and landing places into account. A robot for the field application environment was proposed, and a lightweight global spatial planning technique for the robot based on the graph-search algorithm taking mode switching point optimization into account, with an emphasis on energy efficiency, searching speed, and the viability of real deployment. The fundamental concept is to lower the computational burden by employing an interchangeable search approach that combines planar and spatial search. Furthermore, to safeguard the health of the power battery and the integrity of the mission execution, a trap escape approach was also provided. Simulations are run to test the effectiveness of the suggested model based on the field DEM map. The simulation results show that our technology is capable of producing finished, plausible 3D paths with a high degree of believability. Additionally, the mode-switching point optimization method efficiently identifies additional acceptable places for mode switching, and the improved paths use less time and energy.

cs.RO↗
arXiv2023

Boosting Weakly-Supervised Temporal Action Localization with Text Information

Due to the lack of temporal annotation, current Weakly-supervised Temporal Action Localization (WTAL) methods are generally stuck into over-complete or incomplete localization. In this paper, we aim to leverage the text information to boost WTAL from two aspects, i.e., (a) the discriminative objective to enlarge the inter-class difference, thus reducing the over-complete; (b) the generative objective to enhance the intra-class integrity, thus finding more complete temporal boundaries. For the discriminative objective, we propose a Text-Segment Mining (TSM) mechanism, which constructs a text description based on the action class label, and regards the text as the query to mine all class-related segments. Without the temporal annotation of actions, TSM compares the text query with the entire videos across the dataset to mine the best matching segments while ignoring irrelevant ones. Due to the shared sub-actions in different categories of videos, merely applying TSM is too strict to neglect the semantic-related segments, which results in incomplete localization. We further introduce a generative objective named Video-text Language Completion (VLC), which focuses on all semantic-related segments from videos to complete the text sentence. We achieve the state-of-the-art performance on THUMOS14 and ActivityNet1.3. Surprisingly, we also find our proposed method can be seamlessly applied to existing methods, and improve their performances with a clear margin. The code is available at https://github.com/lgzlIlIlI/Boosting-WTAL.

cs.CV↗
arXiv2023

All-to-key Attention for Arbitrary Style Transfer

Attention-based arbitrary style transfer studies have shown promising performance in synthesizing vivid local style details. They typically use the all-to-all attention mechanism -- each position of content features is fully matched to all positions of style features. However, all-to-all attention tends to generate distorted style patterns and has quadratic complexity, limiting the effectiveness and efficiency of arbitrary style transfer. In this paper, we propose a novel all-to-key attention mechanism -- each position of content features is matched to stable key positions of style features -- that is more in line with the characteristics of style transfer. Specifically, it integrates two newly proposed attention forms: distributed and progressive attention. Distributed attention assigns attention to key style representations that depict the style distribution of local regions; Progressive attention pays attention from coarse-grained regions to fine-grained key positions. The resultant module, dubbed StyA2K, shows extraordinary performance in preserving the semantic structure and rendering consistent style patterns. Qualitative and quantitative comparisons with state-of-the-art methods demonstrate the superior performance of our approach.

cs.CV↗
arXiv2023

Structure, Stability and Superconductivity of N-doped Lutetium Hydrides at kbar Pressures

The structure of the material responsible for the room temperature and near ambient pressure superconductivity reported in an N-doped lutetium hydride [Nature, 615, 244 (2023)] has not been conclusively determined. Herein, density functional theory calculations are performed in an attempt to uncover what it might be. Guided by a range of strategies including crystal structure prediction and modifications of existing structure types, we present an array of Lu-N-H phases that are dynamically stable at experimentally relevant pressures. Although none of the structures found are thermodynamically stable, and none are expected to remain superconducting above 17 K at 10 kbar, a number of metallic compounds with fcc Lu lattices -- as suggested by the experimental X-ray diffraction measurements of the majority phase -- are identified. The system whose calculated equation of states matches best with that measured for the majority phase is fluorite-type LuH2, whose 10 kbar superconducting critical temperature was estimated to be 0.09 K using the Allen-Dynes modified McMillan equation.

cond-mat.supr-con↗
arXiv2023

NAR-Former: Neural Architecture Representation Learning towards Holistic Attributes Prediction

With the wide and deep adoption of deep learning models in real applications, there is an increasing need to model and learn the representations of the neural networks themselves. These models can be used to estimate attributes of different neural network architectures such as the accuracy and latency, without running the actual training or inference tasks. In this paper, we propose a neural architecture representation model that can be used to estimate these attributes holistically. Specifically, we first propose a simple and effective tokenizer to encode both the operation and topology information of a neural network into a single sequence. Then, we design a multi-stage fusion transformer to build a compact vector representation from the converted sequence. For efficient model training, we further propose an information flow consistency augmentation and correspondingly design an architecture consistency loss, which brings more benefits with less augmentation samples compared with previous random augmentation strategies. Experiment results on NAS-Bench-101, NAS-Bench-201, DARTS search space and NNLQP show that our proposed framework can be used to predict the aforementioned latency and accuracy attributes of both cell architectures and whole deep neural networks, and achieves promising performance. Code is available at https://github.com/yuny220/NAR-Former.

cs.CV↗