检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Tang, Shiyu Gong, Ruihao Wang, Yan Liu, Aishan Wang, Jiakai Chen, Xinyun Yu, Fengwei Liu, Xianglong Song, Dawn Yuille, Alan Torr, Philip H.S. Tao, Dacheng The State Key Lab of Software Development Environment Beihang University Beijing100191 China The SenseTime Beijing100191 China The Berkeley Artificial Intelligence Research Lab University of California Berkeley BerkeleyCA94701 United States The Department of Cognitive Science and Computer Science Johns Hopkins University BaltimoreMD21218 United States The Department of Engineering Science University of Oxford OxfordOX1 3PJ United Kingdom The JD Explore Academy Beijing101111 China

Deep neural networks (DNNs) are vulnerable to adversarial noises, which motivates the benchmark of model robustness. Existing benchmarks mainly focus on evaluating defenses, but there are no comprehensive studies of how architecture design and training techniques affect robustness. Comprehensively benchmarking their relationships is beneficial for better understanding and developing robust DNNs. Thus, we propose RobustART, the first comprehensive Robustness investigation benchmark on ImageNet regarding ARchitecture design (49 human-designed off-the-shelf architectures and 1200+ networks from neural architecture search) and Training techniques (10+ techniques, e.g., data augmentation) towards diverse noises (adversarial, natural, and system noises). Extensive experiments substantiated several insights for the first time, e.g., (1) adversarial training is effective for the robustness against all noises types for Transformers and MLP-Mixers;(2) given comparable model sizes and aligned training settings, CNNs > Transformers > MLP-Mixers on robustness against natural and system noises;Transformers > MLP-Mixers > CNNs on adversarial robustness;(3) for some light-weight architectures, increasing model sizes or using extra data cannot improve robustness. Our benchmark http://***/presents: (1) an open-source platform for comprehensive robustness evaluation;(2) a variety of pre-trained models to facilitate robustness evaluation;and (3) a new view to better understand the mechanism towards designing robust DNNs. We will continuously develop to this ecosystem for the community. Copyright © 2021, The Authors. All rights reserved.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Topic Model Based Android Malware Detection 12th

Topic Model Based Android Malware Detection

引用

12th International Conference on Security, Privacy, and Anonymity in Computation, Communication, and Storage, SpaCCS 2019

作者： Song, Yucai Chen, Yang Lang, Bo Liu, Hongyu Chen, Shaojie State Key Lab of Software Development Environment School of Computer Science and Engineering Beihang University Beijing100191 China National Computer Network Emergency Response Technical Team/Coordination Center of China Beijing100191 China

ISBN: (纸本)9783030249069

Nowadays, the security risks brought by Android malwares are increasing. Machine learning is considered as a potential solution for promoting the performance of malware detection. For machine learning based Android malware detection, feature extraction plays a key role. Thinking the source codes of applications are comparable with text documents, we propose a new Android malware detection method based on the topic model which is an effective technique in text feature extraction. Our method regards the decompiled codes of an application as a text document, and the topic model is used to mine the potential topics in the codes which can reflect the semantic feature of the application. The experimental results demonstrate that, our approach performs better than the state-of-the-art methods. Also, our method mines the features in the application files automatically without manually design, and therefore overcomes the limitation in present methods which relies on experts’ prior knowledge. © Springer Nature Switzerland AG 2019.

关键词： Android malware

来源：评论

学校读者我要写书评

暂无评论

Object Detection in Aerial Images: A Large-Scale Benchmark and Challenges

arXiv

引用

arXiv 2021年

作者： Ding, Jian Xue, Nan Xia, Gui-Song Bai, Xiang Yang, Wen Yang, Michael Ying Belongie, Serge Luo, Jiebo Datcu, Mihai Pelillo, Marcello Zhang, Liangpei The State Key Lab LIESMARS Wuhan University Wuhan430079 China The National Engineering Research Center for Multimedia Software School of Computer Science and Institute of Artificial Intelligence Wuhan University Wuhan430072 China The National Engineering Research Center for Multimedia Software School of Computer Science and Institute of Artificial Intelligence The State Key Lab. LIESMARS Wuhan University Wuhan430072 China The School of Electronic Information Huazhong University of Science and Technology Wuhan430079 China The School of Electronic Information The State Key Lab. LIESMARS Wuhan University Wuhan430072 China University of Twente Netherlands Department of Computer Science Cornell University Cornell Tech United States Department of Computer Science University of Rochester RochesterNY14627 United States 82234 Germany Romania DAIS Ca' Foscari University of Venice Italy

In the past decade, object detection has achieved significant progress in natural images but not in aerial images, due to the massive variations in the scale and orientation of objects caused by the bird’s-eye view of aerial images. More importantly, the lack of large-scale benchmarks has become a major obstacle to the development of object detection in aerial images (ODAI). In this paper, we present a large-scale Dataset of Object deTection in Aerial images (DOTA) and comprehensive baselines for ODAI. The proposed DOTA dataset contains 1,793,658 object instances of 18 categories of oriented-bounding-box annotations collected from 11,268 aerial images. Based on this large-scale and well-annotated dataset, we build baselines covering 10 state-of-the-art algorithms with over 70 configurations, where the speed and accuracy performances of each model have been evaluated. Furthermore, we provide a code library for ODAI and build a website for evaluating different algorithms. Previous challenges run on DOTA have attracted more than 1300 teams worldwide. We believe that the expanded large-scale DOTA dataset, the extensive baselines, the code library and the challenges can facilitate the designs of robust algorithms and reproducible research on the problem of object detection in aerial images. Copyright © 2021, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

A comparative study of large-scale cluster workload traces via multiview analysis 21

A comparative study of large-scale cluster workload traces v...

引用

21st IEEE International Conference on High Performance Computing and Communications, 17th IEEE International Conference on Smart City and 5th IEEE International Conference on Data science and Systems, HPCC/SmartCity/DSS 2019

作者： Ruan, Li Xu, Xiangrong Xiao, Limin Yuan, Feng Li, Yin Dai, Dong State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Beijing China Institute of Software Application Technology Guangzhou AND Chinese Academy of Sciences Guangzhou China Department of Computer Science University of North Carolina Charlotte CharlotteNC United States

ISBN: (纸本)9781728120584

Understanding the characteristics of workloads of the large-scale Clusters is the key to diagnose the system bottlenecks, making optimal configuration decisions, improving the system throughput and resource usage. Due to the diversity and multiview of the workload traces, featuring the good designs and bottlenecks by analyzing the workloads under realworld scenarios becomes increasingly challenging. This paper introduces a multiview based trace comparative analysis method by comparatively characterizing how the architecture, jobs, tasks, machines, and resources usage were managed among different platforms. A case study is performed which verified the effectiveness of our method by comparatively analyzing two most representative big traces: Google trace and Alibaba 2018 trace. Quantitative findings, together with the performance bottleneck inferences and suggestions are also presented. To the best of our knowledge, we are the first to perform such comparative empirical study on these two traces using a multiview based approach. Our multifaceted analyses and new findings not only reveal insights that we believe are useful for system designers, IT practitioners and users, but also can promote more researches on big trace data comparative analysis in large-scale clusters. © 2019 IEEE.

关键词： Trace analysis

来源：评论

学校读者我要写书评

暂无评论

Record High Temperatures in the Ocean in 2024

引用

Advances in Atmospheric sciences 2025年第6期42卷 1092-1109页

作者： Lijing CHENG John ABRAHAM Kevin E.TRENBERTH James REAGAN Huai-Min ZHANG Andrea STORTO Karina VON SCHUCKMANN Yuying PAN Yujing ZHU Michael E.MANN Jiang ZHU Fan WANG Fujiang YU Ricardo LOCARNINI John FASULLO Boyin HUANG Garrett GRAHAM Xungang YIN Viktor GOURETSKI Fei ZHENG Yuanlong LI Bin ZHANG Liying WAN Xingrong CHEN Dakui WANG Licheng FENG Xiangzhou SONG Yulong LIU Franco RESEGHETTI Simona SIMONCELLI Gengxin CHEN Rongwang ZHANG Alexey MISHONOV Zhetao TAN Wangxu WEI Huifeng YUAN Guancheng LI Qiuping REN Lijuan CAO Yayang LU Juan DU Kewei LYU Albertus SULAIMAN Michael MAYER Huizan WANG Zhanhong MA Senliang BAO Henqian YAN Zenghong LIU Chunxue YANG Xu LIU Zeke HAUSFATHER Tanguy SZEKELY Flora GUES National Key Laboratory of Earth System Numerical Modeling and Application Institute of Atmospheric PhysicsChinese Academy of SciencesBeijing 100029China University of St.Thomas School of EngineeringMinnesota 55105USA NSF National Center for Atmospheric Research BoulderColorado 80307USA University of Auckland Auckland 0630New Zealand National Oceanic and Atmospheric Administration National Centers for Environmental InformationSilver SpringMaryland 20910USA National Oceanic and Atmospheric Administration National Centers for Environmental InformationAshevilleNC 28801USA National Research Council(CNR)Institute of Marine Sciences(ISMAR) Rome 00133Italy Mercator Ocean International Toulouse 31400France Department of Earth and Environmental Science University of PennsylvaniaPhiladelphiaPennsylvania 19104USA Institute of Oceanology Chinese Academy of SciencesQingdao 266071China National Marine Environmental Forecasting Center Ministry of Natural Resources of ChinaBeijing 100081China North Carolina Institute for Climate Studies(NCICS) North Carolina State UniversityAshevilleNC 28804USA Oceanographic Data Center Chinese Academy of SciencesQingdao 266071China College of Oceanography Hohai UniversityNanjing 210098China National Marine Data and Information Service Tianjin 300171China Istituto Nazionale di Geofisica e Vulcanologia Sede di BolognaBologna 40128Italy South China Sea Institute of Oceanology Chinese Academy of SciencesGuangzhou 510301China ESSIC/CISESS-MD University of MarylandCollege ParkMD 20740USA Computer Network Information Center Chinese Academy of SciencesBeijing 100083China Eco-Environmental Monitoring and Research Center Pearl River Valley and South China Sea Ecology and Environment AdministrationMinistry of Ecology and EnvironmentGuangzhou 510611China National Meteorological Information Center China Meteorological AdministrationBeijing 100081China International Research Center of Big Data for Sustainable Development Goals Beijing 100094China Xiamen Universit

Heating in the ocean has continued in 2024 in response to increased greenhouse gas concentrations in the atmosphere,despite the transition from an El Ni?o to neutral conditions. In 2024, both global sea surface temperature(SST) and upper2000 m ocean heat content(OHC) reached unprecedented highs in the historical record. The 0–2000 m OHC in 2024exceeded that of 2023 by 16 ± 8 ZJ(1 Zetta Joules = 1021 Joules, with a 95% confidence interval)(IAP/CAS data), which is confirmed by two other data products: 18 ± 7 ZJ(CIGAR-RT reanalysis data) and 40 ± 31 ZJ(Copernicus Marine data,updated to November 2024). The Indian Ocean, tropical Atlantic, Mediterranean Sea, North Atlantic, North Pacific, and Southern Ocean also experienced record-high OHC values in 2024. The global SST continued its record-high values from2023 into the first half of 2024, and declined slightly in the second half of 2024, resulting in an annual mean of 0.61°C ±0.02°C(IAP/CAS data) above the 1981–2010 baseline, slightly higher than the 2023 annual-mean value(by 0.07°C ±0.02°C for IAP/CAS, 0.05°C ± 0.02°C for NOAA/NCEI, and 0.06°C ± 0.11°C for Copernicus Marine). The record-high values of 2024 SST and OHC continue to indicate unabated trends of global heating.

关键词： ocean heat content sea surface temperature ocean temperature global warming climate

来源：评论

学校读者我要写书评

暂无评论

Interpretable Spatiotemporal Deep Learning Model for Traffic Flow Prediction Based on Potential Energy Fields

Interpretable Spatiotemporal Deep Learning Model for Traffic...

引用

IEEE International Conference on Data Mining (ICDM)

作者： Jiahao Ji Jingyuan Wang Zhe Jiang Jingtian Ma Hu Zhang School of Computer Science and Engineering Beihang University Beijing China State Key Laboratory of Software Development Environment Beihang University Beijing China MOE Engineering Research Center of ACAT Beihang University Beijing China Department of Computer Science The University of Alabama Tuscaloosa Alabama USA

ISBN: (数字)9781728183169

ISBN: (纸本)9781728183176

Traffic flow prediction is of great importance in traffic management and public safety, but is challenging due to the complex spatial-temporal dependencies as well as temporal dynamics. Existing work either focuses on traditional statistical models, which have limited prediction accuracy, or relies on black-box deep learning models, which have superior prediction accuracy but are hard to interpret. In contrast, we propose a novel interpretable spatiotemporal deep learning model for traffic flow prediction. Our main idea is to model the physics of traffic flow through a number of latent Spatio-Temporal Potential Energy Fields (ST-PEFs), similar to water flow driven by the gravity field. We develop a Wind field Decomposition (WD) algorithm to decompose traffic flow into poly-tree components so that ST-PEFs can be established. We then design a spatiotemporal deep learning model for the ST-PEFs, which consists of a temporal component (modeling the temporal correlation) and a spatial component (modeling the spatial dependencies). To the best of our knowledge, this is the first work that make traffic flow prediction based on ST-PEFs. Experimental results on real-world traffic datasets show the effectiveness of our model compared to the existing methods. A case study confirms our model interpretability.

关键词： Deep learning Potential energy Predictive models Prediction algorithms Data models Spatiotemporal phenomena Safety

来源：评论

学校读者我要写书评

暂无评论

On PID control for synchronization of complex dynamical network with delayed nodes

引用

science China(Technological sciences) 2019年第8期62卷 1412-1422页

作者： GU HaiBo Lv JinHu LIN ZongLi Key Laboratory of Systems and Control Academy of Mathematics and Systems Science Chinese Academy of Sciences Beijing 100190 China School of Mathematical Sciences University of Chinese Academy of Sciences Beijing 100049 China School of Automation Science and Electrical Engineering State Key Laboratory of Software Development Environment and Beijing Advanced Innovation Center for Big Data and Brain Machine Intelligence Beihang University Beijing 100083 China Charles L. Brown Department of Electrical and Computer Engineering University of Virginia Charlottesville VA 22904-4743 USA

Over the past two decades, synchronization, as an interesting collective behavior of complex dynamical networks, has been attracting much attention. To reveal and analyze the inherent mechanism of synchronization in complex dynamical networks with time delays in nodes, this paper attempts to use PD and PI control protocols to achieve synchronization. Based on a classical network model, we investigate the PD and PI control for synchronization of complex dynamical networks with delayed nodes and obtain some sufficient conditions. By using Lyapunov functions and appropriate state transformations, we prove that global synchronization can be achieved via the above control protocols. Finally, some simulation examples are illustrated to validate the effectiveness of the proposed theoretical results.

关键词： synchronization PID control complex dynamical network networked control system delayed node

来源：评论

学校读者我要写书评

暂无评论

DeePMD-kit v3: A Multiple-Backend Framework for Machine Learning Potentials

arXiv

引用

arXiv 2025年

作者： Zeng, Jinzhe Zhang, Duo Peng, Anyang Zhang, Xiangyu He, Sensen Wang, Yan Liu, Xinzijian Bi, Hangrui Li, Yifan Cai, Chun Zhang, Chengqian Du, Yiming Zhu, Jia-Xin Mo, Pinghui Huang, Zhengtao Zeng, Qiyu Shi, Shaochen Qin, Xuejian Yu, Zhaoxi Luo, Chenxing Ding, Ye Liu, Yun-Pei Shi, Ruosong Wang, Zhenyu Bore, Sigbjørn Løland Chang, Junhan Deng, Zhe Ding, Zhaohan Han, Siyuan Jiang, Wanrun Ke, Guolin Liu, Zhaoqing Lu, Denghui Muraoka, Koki Oliaei, Hananeh Singh, Anurag Kumar Que, Haohui Xu, Weihong Xu, Zhangmancang Zhuang, Yong-Bin Dai, Jiayu Giese, Timothy J. Jia, Weile Xu, Ben York, Darrin M. Zhang, Linfeng Wang, Han School of Artificial Intelligence and Data Science Unversity of Science and Technology of China Hefei China AI for Science Institute Beijing100080 China DP Technology Beijing100080 China Academy for Advanced Interdisciplinary Studies Peking University Beijing100871 China State Key Lab of Processors Institute of Computing Technology Chinese Academy of Sciences Beijing100871 China University of Chinese Academy of Sciences Beijing China Baidu Inc. Beijing China Department of Computer Science University of Toronto TorontoON Canada Department of Chemistry Princeton University PrincetonNJ08540 United States University of Chinese Academy of Sciences Beijing100871 China State Key Laboratory of Physical Chemistry of Solid Surfaces iChEM College of Chemistry and Chemical Engineering Xiamen University Xiamen361005 China College of Integrated Circuits Hunan University Changsha410082 China State Key Laboratory of Advanced Technology for Materials Synthesis and Processing Center for Smart Materials and Device Integration School of Material Science and Engineering Wuhan University of Technology Wuhan430070 China College of Science National University of Defense Technology Changsha410073 China Hunan Key Laboratory of Extreme Matter and Applications National University of Defense Technology Changsha410073 China ByteDance Research Beijing100098 China Ningbo Institute of Materials Technology and Engineering Chinese Academy of Sciences Ningbo315201 China College of Materials Science and Opto-Electronic Technology University of Chinese Academy of Sciences Beijing100049 China Key Laboratory of Theoretical and Computational Photochemistry of Ministry of Education College of Chemistry Beijing Normal University Beijing100875 China Department of Geosciences Princeton University PrincetonNJ08544 United States Department of Applied Physics and Applied Mathematics Columbia University New YorkNY10027 United States IKKEM Fujian Xiamen361005 China Graduate

In recent years, machine learning potentials (MLPs) have become indispensable tools in physics, chemistry, and materials science, driving the development of software packages for molecular dynamics (MD) simulations and related applications. These packages, typically built on specific machine learning frameworks such as TensorFlow, PyTorch, or JAX, face integration challenges when advanced applications demand communication across different frameworks. The previous TensorFlow-based implementation of DeePMD-kit exemplified these limitations. In this work, we introduce DeePMD-kit version 3, a significant update featuring a multi-backend framework that supports TensorFlow, PyTorch, JAX, and PaddlePaddle backends, and demonstrate the versatility of this architecture through the integration of other MLPs packages and of Differentiable Molecular Force Field. This architecture allows seamless backend switching with minimal modifications, enabling users and developers to integrate DeePMD-kit with other packages using different machine learning frameworks. This innovation facilitates the development of more complex and interoperable workflows, paving the way for broader applications of MLPs in scientific research. Copyright © 2025, The Authors. All rights reserved.

关键词： Molecular dynamics

来源：评论

学校读者我要写书评

暂无评论

A Learning Analytic Model for Smart Classroom

引用

Asia Pacific Web (APWeb) and Web-Age Information Management (WAIM) Joint Conference on Web and Big Data, APWeb-WAIM 2018

作者： Wang, Qunbo Wu, Wenjun Qi, Yuxing State Key Lab of Software Development Environment Department of Computer Science and Engineering Beihang University Beijing China

ISBN: (纸本)9783030012977

With the popularity of Smart Classroom, it is necessary to study corresponding learning analytic methods to assist instructors. However, little research has investigated analyzing hidden state in class, which is an important analysis work. Therefore, focusing on the interactive learning through individual Pad devices, we propose a Learning Analytic Model to analyze hidden state with students’ sequential behaviors that automatically recorded by devices. The model segments the class’ process into multiple phases and construct a Hidden Markov Model (HMM) to infer students’ state. In addition, a web page is developed to show students’ behaviors and related analysis results intuitively. The experiment shows our model can fine-grained analyze and feedback the learning state of students in the smart classroom, which effectively help instructors improve teaching methods. © 2018, Springer Nature Switzerland AG.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

A Learning Analytics System for Cognition Analysis in Online Learning Community

引用

Asia Pacific Web (APWeb) and Web-Age Information Management (WAIM) Joint Conference on Web and Big Data, APWeb-WAIM 2018

作者： Wu, Yinan Wu, Wenjun State Key Lab of Software Development Environment Department of Computer Science and Engineering Beihang University Beijing China

ISBN: (纸本)9783030012977

While cognitive behaviors and social network structure in Online Learning Community (OLC) have been studied in the past, few research has proposed a model linking the two important factors to analyze students’ cognitive learning gains, even though it has been widely acknowledged that interaction is a significant way for students to exchange knowledge and obtain learning gains. In this paper, for a better indication of cognitive gains, we introduce an analytic model to quantify the students’ learning gains by using a redesigned taxonomy of cognitive behaviors while considering the flow of knowledge among students in discussion forums. And further, we implement a learning analytics system to streamline the data analysis pipeline of social network analysis, cognition classification and learning gain calculation and visualize the analytic results from multiple-level views including student, discussion thread and forum. We demonstrate the results on a MOOC course and confirm the effectiveness of our model. Our model and analytic system enable instructors and TAs to take active mediation among online discussions of students to improve their cognitive gains through OLCs. © 2018, Springer Nature Switzerland AG.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：