检索结果-内蒙古大学图书馆

arXiv 2019年

作者： Yang, Hongyu Huang, Di Wang, Yunhong Jain, Anil K. Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing100191 China Department of Computer Science and Engineering Michigan State University East LansingMI48824 United States

The two underlying requirements of face age progression, i.e. aging accuracy and identity permanence, are not well studied in the literature. This paper presents a novel generative adversarial network based approach to address the issues in a coupled manner. It separately models the constraints for the intrinsic subject-specific characteristics and the age-specific facial changes with respect to the elapsed time, ensuring that the generated faces present desired aging effects while simultaneously keeping personalized properties stable. To ensure photo-realistic facial details, high-level age-specific features conveyed by the synthesized face are estimated by a pyramidal adversarial discriminator at multiple scales, which simulates the aging effects with finer details. Further, an adversarial learning scheme is introduced to simultaneously train a single generator and multiple parallel discriminators, resulting in smooth continuous face aging sequences. The proposed method is applicable even in the presence of variations in pose, expression, makeup, etc., achieving remarkably vivid aging effects. Quantitative evaluations by a COTS face recognition system demonstrate that the target age distributions are accurately recovered, and 99.88% and 99.98% age progressed faces can be correctly verified at 0.001% FAR after age transformations of approximately 28 and 23 years elapsed time on the MORPH and CACD databases, respectively. Both visual and quantitative assessments show that the approach advances the state-of-the-art. Copyright © 2019, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Field-free spin-orbit torque-induced switching of perpendicular magnetization in a ferrimagnetic layer with vertical composition gradient

arXiv

引用

arXiv 2021年

作者： Zheng, Zhenyi Zhang, Yue Lopez-Dominguez, Victor Sánchez-Tejerina, Luis Shi, Jiacheng Feng, Xueqiang Chen, Lei Wang, Zilu Zhang, Zhizhong Zhang, Kun Hong, Bin Xu, Yong Zhang, Youguang Carpentieri, Mario Fert, Albert Finocchio, Giovanni Zhao, Weisheng Amiri, Pedram Khalili Fert Beijing Research Institute School of Integrated Circuit Science and Engineering Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing China Department of Electrical and Computer Engineering Northwestern University EvanstonIL United States School of Electronics and Information Engineering Beihang University Beijing China Department of Mathematical and Computer Sciences Physical Sciences and Earth Sciences University of Messina Messina Italy Dipartimento di Ingegneria Elettrica e dell’Informazione Politecnico di Bari Bari Italy Unité Mixte de Physique CNRS Thales Univ. Paris-Sud University of Paris-Saclay Palaiseau France

Current-induced spin-orbit torques (SOTs) are of interest for fast and energy-efficient manipulation of magnetic order in spintronic devices. To be deterministic, however, switching of perpendicularly magnetized materials by SOT requires a mechanism for in-plane symmetry breaking. Existing methods to do so involve the application of an in-plane bias magnetic field, or incorporation of in-plane structural asymmetry in the device, both of which can be difficult to implement in practical applications. Here, we report bias-field-free SOT switching in a single perpendicular CoTb layer with an engineered vertical composition gradient. The vertical structural inversion asymmetry induces strong intrinsic SOTs and a gradient-driven Dzyaloshinskii–Moriya interaction (g-DMI), which breaks the in-plane symmetry during the switching process. Micromagnetic simulations are in agreement with experimental results, and elucidate the role of g-DMI in the deterministic switching. This bias-field-free switching scheme for perpendicular ferrimagnets with g-DMI provides a strategy for efficient and compact SOT device design. © 2021, CC BY.

关键词： Switching

来源：评论

学校读者我要写书评

暂无评论

Application Of LSTM In Protein Structure Prediction LINA

Application Of LSTM In Protein Structure Prediction LINA

引用

International Conference on Wavelet Analysis and Pattern Recognition (ICWAPR)

作者： Lina Yang Pu Wei Xichun Li Yuan Yan Tang School of Computer Electronics and Information Guangxi University Nanning China Guangxi Normal University for Nationalities Chongzuo China Beihang University Beijing Advanced Innovation Center for Big Data and Brain Computing (BDBC) Beijing China

In this paper the authors discuss the applications of LSTM Neural Network in Protein Structure Prediction. The main idea is to construct a LSTM neural network. Predicting the secondary structure of a protein is the basis content for predicting its spatial structure. In this article, a position-specific scoring matrices (PSSM) containing evolutionary information is linked to other features to construct a completely new feature set. The CB513 data set is selected to construct LSTM neural networks to predict the secondary structure of the sequence. Experiments have shown that the proposed method effectively improves the prediction accuracy and is better than the previous method. The idea in this paper can also be applied to the analysis of other sequences.

关键词： Biological neural networks Proteins Neurons Logic gates Biological information theory Amino acids

来源：评论

学校读者我要写书评

暂无评论

A One-step Pruning-recovery Framework for Acceleration of Convolutional Neural Networks

A One-step Pruning-recovery Framework for Acceleration of Co...

引用

International Conference on Tools for Artificial Intelligence (ICTAI)

作者： Dong Wang Xiao Bai Lei Zhou Jun Zhou School of Computer Science and Engineering Beijing Advanced Innovation Center for Big Data and Brain Computing Jiangxi Research Institute Beihang University Beijing China School of Information and Communication Technology Griffith University Queensland Australia

Acceleration of convolutional neural network has received increasing attention during the past several years. Among various acceleration techniques, filter pruning has its inherent merit by effectively reducing the number of convolution filters. However, most filter pruning methods resort to tedious and time-consuming layer-by-layer pruning-recovery strategy to avoid a significant drop of accuracy. In this paper, we present an efficient filter pruning framework to solve this problem. Our method accelerates the network in one-step pruning-recovery manner with a novel optimization objective function, which achieves higher accuracy with much less cost compared with existing pruning methods. Furthermore, our method allows network compression with global filter pruning. Given a global pruning rate, it can adaptively determine the pruning rate for each single convolutional layer, while these rates are often set as hyper-parameters in previous approaches. Evaluated on VGG- 16 and ResNet-50 using ImageNet, our approach outperforms several state-of-the-art methods with less accuracy drop under the same and even much fewer floating-point operations (FLOPs).

关键词：

来源：评论

学校读者我要写书评

暂无评论

Continuous Control with Deep Reinforcement Learning for Mobile Robot Navigation

Continuous Control with Deep Reinforcement Learning for Mobi...

引用

Chinese Automation Congress (CAC)

作者： Jiaqi Xiang Qingdong Li Xiwang Dong Zhang Ren School of Automation Science and Electrical Engineering Beihang University Beijing P.R. China Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing P.R. China

Autonomous navigation is one of the focuses in the field of mobile robot research. The traditional method usually consists of two parts: building the map of environment, localization of mobile robot and path planning. However, these traditional methods usually rely on high-precision sensor information. At the same time, mobile robots have no intelligent understanding of autonomous navigation. In this article, a deep reinforcement learning method, i.e. soft actor critic, is used to navigate in a mapless environment. It takes laser scanning data and information of the target as input, outputs linear velocity and angular velocity in continuous space. The simulation shows that this learning-based end-to-end autonomous navigation method can accomplish tasks as well as traditional methods.

关键词： Mobile robots Machine learning Navigation Autonomous robots Entropy Lasers

来源：评论

学校读者我要写书评

暂无评论

Distributed State Estimation for Heterogeneous Mobile Sensor Networks with Sensor Faults

Distributed State Estimation for Heterogeneous Mobile Sensor...

引用

第三十八届中国控制会议

作者： Yingrong Yu Siting Peng Qingdong Li Xiwang Dong Zhang Ren Science and Technology on Aircraft Control Laboratory School of Automation Science and Electrical Engineering Beihang University Beijing Electro-Mechanical Engineering Institute Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University

Consensus-based distributed state estimation for homogeneous mobile sensor networks has been widely studied currently. However, the sensors used for target detection and tracking in practical applications usually have different physical properties, different measuring range or accuracy. In order to realize data exchange between sensors which are of heterogeneity and to obtain a consistent and precise target state estimation result, an improved Cubature Kalman Filtering(CKF) method based on weighted average consensus theory is put forward herein. Moreover, the adverse impacts caused by sensor faults have been taken into consideration in modeling. The stochastic boundedness of the estimation errors for the discussed sensor networks is proved in the paper. Finally, numerical simulations are provided to demonstrate the validity of the proposed method.

关键词： Cubature Kalman Filter weighted average consensus theory target state estimation mobile sensor networks stochastic boundedness

来源：评论

学校读者我要写书评

暂无评论

KE-GAN: Knowledge Embedded Generative Adversarial Networks for Semi-Supervised Scene Parsing

KE-GAN: Knowledge Embedded Generative Adversarial Networks f...

引用

IEEE/CVF Conference on Computer Vision and Pattern Recognition

作者： Mengshi Qi Yunhong Wang Jie Qin Annan Li State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University China Beijing Advanced Innovation Center for Big Data and Brain Computing Inception Institute of Artificial Intelligence UAE

In recent years, scene parsing has captured increasing attention in computer vision. Previous works have demonstrated promising performance in this task. However, they mainly utilize holistic features, whilst neglecting the rich semantic knowledge and inter-object relationships in the scene. In addition, these methods usually require a large number of pixel-level annotations, which is too expensive in practice. In this paper, we propose a novel Knowledge Embedded Generative Adversarial Networks, dubbed as KE-GAN, to tackle the challenging problem in a semi-supervised fashion. KE-GAN captures semantic consistencies of different categories by devising a Knowledge Graph from the large-scale text corpus. In addition to readily-available unlabeled data, we generate synthetic images to unveil rich structural information underlying the images. Moreover, a pyramid architecture is incorporated into the discriminator to acquire multi-scale contextual information for better parsing results. Extensive experimental results on four standard benchmarks demonstrate that KE-GAN is capable of improving semantic consistencies and learning better representations for scene parsing, resulting in the state-of-the-art performance.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Embedded Partitioning Real-Time Operating System Based on Microkernel

Embedded Partitioning Real-Time Operating System Based on Mi...

引用

IEEE International Conference on Computational Science and Engineering, CSE

作者： Tanhong Chen Huiyong Li Jianwei Niu Tao Ren Guizhou Xu School of Computer Science Beihang University Hangzhou Innovation Institute Beihang University Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University BEIJING ACOINFO TECHNOLOGY CO. LTD.

The embedded partitioning operating system is active in security-related fields such as avionics and automotive embedded software. The temporal and spatial isolation of the partition is the guarantee of system security. In addition, as these areas become more demanding for applications, improving the real-time performance of partitioning operating systems has become a trend. To reach this goal, we propose a solution that leverages the microkernel architecture in the embedded partitioning operating system. The microkernel ensures the determinacy of real-time scheduling policies and system calls by simplifying kernel functionality and preserving the most basic resources. Partitions built on top of the microkernel enable low-cost communication, fast context switching and mixed task scheduling. Based on the microkernel, we developed a complete real-time partitioning operating system. The system supports multi-core processors to take full advantage of the high performance of existing embedded processors. On different hardware platforms, we conducted two types of tests about interrupt response time and latency. The results show that our proposed architecture exhibits high real-time performance.

关键词： Task analysis Real-time systems Kernel Virtual machine monitors Security

来源：评论

学校读者我要写书评

暂无评论

Transferring knowledge distillation for multilingual social event detection

arXiv

引用

arXiv 2021年

作者： Ren, Jiaqian Peng, Hao Jiang, Lei Wu, Jia Tong, Yongxin Wang, Lihong Bai, Xu Wang, Bo Yang, Qiang Institute of Information Engineering Chinese Academy of Sciences The School of Cyber Security University of Chinese Academy of Sciences Beijing100041 China Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing100083 China The School of Cyber Science and Technology Beihang University Beijing100083 China Institute of Information Engineering Chinese Academy of Sciences Beijing100041 China The Department of Computing Macquarie University Sydney Australia The State Key Laboratory of Software Development Environment The School of Computer Science and Engineering Beihang University Beijing100083 China The National Computer Network Emergency Response Technical Team Coordination Center of China Beijing100029 China The Department of Computer Science and Engineering Hong Kong University of Science and Technology Hong Kong AI Group WeBank Co. Ltd. China

Recently published graph neural networks (GNNs) show promising performance at social event detection tasks. However, most studies are oriented toward monolingual data in languages with abundant training samples. This has left the more common multilingual settings and lesser-spoken languages relatively unexplored. Thus, we present a GNN that incorporates cross-lingual word embeddings for detecting events in multilingual data streams. The first exploit is to make the GNN work with multilingual data. For this, we outline a construction strategy that aligns messages in different languages at both the node and semantic levels. Relationships between messages are established by merging entities that are the same but are referred to in different languages. Non-English message representations are converted into English semantic space via the cross-lingual word embeddings. The resulting message graph is then uniformly encoded by a GNN model. In special cases where a lesser-spoken language needs to be detected, a novel cross-lingual knowledge distillation framework, called CLKD, exploits prior knowledge learned from similar threads in English to make up for the paucity of annotated data. Experiments on both synthetic and real-world datasets show the framework to be highly effective at detection in both multilingual data and in languages where training samples are scarce. © 2021, CC BY.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

Generating domain ontology from Chinese customer reviews to analysis fine-gained product quality risk 2018

Generating domain ontology from Chinese customer reviews to ...

引用

2018 International Conference on computing and data Engineering, ICCDE 2018

作者： Lin, Shuo Han, Jun Kumar, Kuldeep Wang, Jiping Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University 100191 China Beihang University 37 Xueyuan Road Beijing China Bond University QLD4229 Australia

ISBN: (纸本)9781450363938

With the rapid development of E-commerce in China, quality of the products on online shopping platforms has caused wide concern. Customer reviews, which commented by people who bought the very product, now have been one of the most important resources for analyzing product’s quality risk. We can get fine-gained, aspect-oriented risk information of a product by mining its reviews. Unfortunately, people tend to write reviews with casual grammar or just omit parts of components of a sentence. Both these features will cause negative impacts when parsing the raw customer reviews directly. Thus a knowledge base which is built totally beyond the reviews could be used to analyze it despite the drawbacks above. In this paper, we generate a domain ontology from raw text in the online encyclopedia. It can be viewed as a graph whose nodes represent domain concepts and edges represent the relations between these concepts. In our work, we integrate syntactic tree structure in linear-chain CRFs for recognizing domain concepts and train SVMs and MaxEnt models on elaborate features for clarifying three types of relationship, namely "Attribute-of", "Part-of" and "Instance-of". Once the ontology has been built, product properties with potential risk will be extracted by our matching method. Experiment show that our approach achieves 64.4% precision and 82.4% recall on risky property extraction task. © 2018 Association for computing Machinery.

关键词： Sentiment analysis

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：