检索结果-内蒙古大学图书馆

5th International Academic Exchange Conference on science and Technology Innovation, IAECST 2023

作者： Liu, Jiaxing Liu, Chengyu Beijing University of Posts and Telecommunications School of Computer Science National Pilot Software Engineering School Beijing China

ISBN: (纸本)9798350357738

With the extensive application of deep learning, neural network-based Visual Inertial Odometry (VIO) has demonstrated significantly better performance than traditional geometric VIO. Deep learning-based VIO fuses information from vision and inertial measurement unit (IMU) to learn positional changes in an end-to-end data-driven manner. However, existing studies have generally focused only on the extraction of representations from raw data, ignoring the correlations and interactions between modalities. Moreover, these end-to-end networks are often criticized for their limited interpretability. This study proposes an innovative deep learning VIO model that incorporates mechanical choreography. The model enhances the network with prior positional information through the integration of results from IMU mechanical choreography. It leverages multimodal feature learning concepts and employs a cross-modal co-attention mechanism to effectively merge features extracted from images and IMUs. This approach aims to achieve highly accurate, efficient, and interpretable position estimation tasks. Finally, the trained model is tested on a public dataset and the accuracy is compared with traditional methods and classical neural VIO to verify the effectiveness and feasibility of the proposed method in this study. © 2023 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Band selection based hyperspectral unmixing

Band selection based hyperspectral unmixing

引用

2009 IEEE International Workshop on Imaging Systems and Techniques, IST 2009

作者： Sen, Jia Zhen, Ji Yuntao, Qian Texas Instruments DSPs Lab College of Computer Science and Software Engineering Shenzhen University Shenzhen 518060 China College of Computer Science Zhejiang University Hangzhou China

ISBN: (纸本)9781424434831

Hyperspectral unmixing is the procedure by which the measured spectrum of a mixed pixel is decomposed into a collection of constituent spectra, or endmembers, and their mixing proportions. However, due to the hundreds of spectral bands contained in the hyperspectral imagery, the large amount of data not only increase the computational loads, but also are unfavorable for the fast hyperspectral unmixing. Hence, dimensionality reduction, which selects the relevant range of wavelengths in the spectrum, is a necessary preprocessing step for hyperspectral unmixing. In this paper, in order to preserve the crucial and critical information, band selection techniques are firstly used to choose the appropriate bands from the original data, and then the unmixing methods are applied. Two recently proposed algorithms, affinity propagation and constrained nonnegative matrix factorization, are respectively adopted for the above two procedures. Experimental results show that the performance of the band selection based hyperspectral unmixing strategy is comparable to that without band selection. © 2009 IEEE.

关键词： Matrix algebra

来源：评论

学校读者我要写书评

暂无评论

Distributed Secure Storage Scheme Based on Sharding Blockchain

引用

computers, Materials & Continua 2022年第3期70卷 4485-4502页

作者： Jin Wang Chenchen Han Xiaofeng Yu Yongjun Ren R.Simon Sherratt School of Computer Science and Mathematics Fujian University of TechnologyFuzhou350118China School of Computer&Communication Engineering Changsha University of Science&TechnologyChangsha410004China School of Business Nanjing UniversityNanjing210093China School of Computer School of SoftwareSchool of Cyberspace SecurityEngineering Research Center of Digital ForensicsMinistry of EducationNanjing University of Information Science&TechnologyNanjing210044China Department of Biomedical Engineering University of ReadingRG66AYUK

Distributed storage can store data in multiple devices or servers to improve data ***,in today’s explosive growth of network data,traditional distributed storage scheme is faced with some severe challenges such as insufficient performance,data tampering,and data lose.A distributed storage scheme based on blockchain has been proposed to improve security and efficiency of traditional distributed *** this scheme,the following improvements have been made in this *** paper first analyzes the problems faced by distributed *** proposed to build a new distributed storage blockchain scheme with sharding *** proposed scheme realizes the partitioning of the network and nodes by means of blockchain sharding technology,which can improve the efficiency of data verification between *** addition,this paper uses polynomial commitment to construct a new verifiable secret share scheme called *** new scheme is one of the foundations for building our improved distributed storage blockchain *** with the previous scheme,our new scheme does not require a trusted third party and has some new features such as homomorphic and batch *** security of VSS can be further *** comparisons show that the proposed scheme significantly reduces storage and communication costs.

关键词： Blockchain distributed storage verifiable secret share polynomial commitment

来源：评论

学校读者我要写书评

暂无评论

Analyzing & enhancing energy efficient communication protocol for wireless micro-sensor networks

Analyzing & enhancing energy efficient communication protoco...

引用

1st International Conference on Information and Communication Technology, ICICT 2005

作者： Butt, Irfan Majeed Khan, Shoab Ahmed Computer Software Engineering National University of Science and Technology Pakistan Tamiz-ud-Din Road Rawalpindi Pakistan

ISBN: (纸本)0780394216

The paper presents an analysis of energy efficient routing protocols with direct communication protocol. A comparison of these protocols is made analyzing energy consumption at each node and explaining system lifetime after certain rounds. The paper also proposes a novel energy conscious cluster head selection algorithm for making system more reliable and efficient. Simulation shows that our proposed algorithm enhances the system reliability and accuracy. © 2005 IEEE.

关键词： Network protocols

来源：评论

学校读者我要写书评

暂无评论

FSRF:An Improved Random Forest for Classification

FSRF:An Improved Random Forest for Classification

引用

2020 IEEE International Conference on Advances in Electrical engineering and computer Applications, AEECA 2020

作者： Feng, Wenxian Ma, Chenkai Zhao, Guozhang Zhang, Rui Jilin University Department of Software Engineering Changchun China Jilin University Department of Computer Science Changchun China

ISBN: (纸本)9781728165202

Random forest algorithm is a flexible and easy-to-use machine learning algorithm, which is widely used in classification problems. However, the traditional random forest has some limitations. Because the randomness added by random forest to decision trees almost only occurs in the feature selection when the decision trees are generated, the fixity of decision trees generation rules will lead to relatively serious over fitting. At the same time, in the face of data with high and unbalanced feature dimensions, the performance of algorithm is seriously weakened because high-dimensional data usually contains many irrelevant and redundant features. To solve these problems, we propose an improved random forest algorithm FSRF. Based on the traditional random forest algorithm, we use the feature selection methods to preprocess the data and get the feature subset with the best classification performance to construct the random forest. At the same time, we introduce sparse matrix projection to improve the generation of the random forest. Experiments show that our method reduces the influence of redundant features on classification and improves the accuracy. © 2020 IEEE.

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

Knowledge discovery in sportsfinder: An agent to extract sports results from the web 3rd

Knowledge discovery in sportsfinder: An agent to extract spo...

引用

3rd Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 1999

作者： Lu, Hongen Sterling, Leon Wyatt, Alex Intelligent Agent Laboratory Department of Computer Science and Software Engineering University of Melbourne ParkvilleVIC3052 Australia

ISBN: (纸本)3540658661

There is a wealth of information to be mined from the World Wide Web. Unfortunately, standard natural language processing (NLP) extraction techniques perform poorly on the choppy, semi-structured information fragments, such as sports results, which are popular to be published on the Web pages nowadays. In this paper, we present an information agent: SportsFinder, an agent to extract sports scores from the World Wide Web, as well as the knowledge discovering method to learn new express patterns to improve the agent’s performance. © Springer-Verlag Berlin Heidelberg 1999.

关键词： Sports

来源：评论

学校读者我要写书评

暂无评论

Subspace Inference in SDE-Net for Bayesian Deep Neural Networks

Subspace Inference in SDE-Net for Bayesian Deep Neural Netwo...

引用

2021 China Automation Congress, CAC 2021

作者： Wang, YongGuang Yao, ShuZhen Tan, HuoBin School of Computer Science and Engineering Beihang University Beijing China School of Software Beihang University Beijing China

ISBN: (纸本)9781665426473

Bayesian inference was once a gold standard for uncertainty estimation with neural networks. However, Bayesian inference is inefficient in high-dimensional parameter spaces of Deep Neural Networks (DNNs). A novel Non-Bayesian Neural Stochastic Differential Equation (SDE-Net) model quantifies epistemic uncertainty of DNNs from the perspective of dynamical systems. In this paper, we propose a subspace inference procedure in Non-Bayesian SDE-Net method for uncertainty estimation from Bayesian perspectives. This procedure contains two steps, first of all, the low-dimensional parameter subspaces of SDE-Net model are constructed to generate the first principal components from Stochastic Gradient Descent (SGD) trajectories, which involve various high-performance models. Secondly, Variational Inference (VI) and Elliptical Slice Sampling (ESS) methods are implemented in the constructed subspace to explore in the full parameter spaces for uncertainty estimation. The experimental results of Bayesian average on the derived posterior in the generated principal subspaces of model parameters show that accurate and well-calibrated results can be obtained for regression and classification tasks. © 2021 IEEE

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Named Entity Recognition in Emergency Domain based on BERT-BILSTM-CRF 2

Named Entity Recognition in Emergency Domain based on BERT-B...

引用

2nd IEEE International Conference on Electronic Technology, Communication and Information, ICETCI 2022

作者： Tian, Yu Beijing University of Posts and Telecommunications School of Computer Science National Pilot Software Engineering School Beijing China

ISBN: (数字)9781728181158

ISBN: (纸本)9781728181158

Chinese named entity recognition is a key problem in natural language processing. The traditional language processing model can not effectively represent the context semantic information in the text, and can not deal with the different contexts of polysemy, which affects the effect of entity recognition. A BERT-BILSTM-CRF entity recognition model for the emergency domain is proposed. BERT pre-training language model generates word vectors which represent the semantic information of context, and the generated vectors are characterized by two-way long-term and short-term memory network. Self-attention mechanism can effectively obtain the long-distance dependence in text sentences. And finally, decoding is carried out through the CRF to generate an entity label sequence. The experimental results show that the model have achieved excellent results in the emergency domain corpus. © 2022 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Beyond utility and usability: Towards affectability in agile software requirements engineering

Beyond utility and usability: Towards affectability in agile...

引用

2018 International Conference on Computational science and Computational Intelligence, CSCI 2018

作者： Kamthan, Pankaj Shahmir, Nazlie Department of Computer Science and Software Engineering Concordia University Montreal Canada WestJet Airlines Limited Calgary Canada

ISBN: (纸本)9781728113609

As the relationship between users and interactive software systems becomes increasingly personal, the users are expecting more from these systems than just utility and usability. To that effect, in the context of agile development of interactive software systems, a framework for a user-centered and conceptual model-based process for engineering affective user stories is proposed, described, and illustrated. The implications of affect for market-oriented software development are emphasized, and challenges and limitations related to current state of requirements engineering education are discussed. © 2018 IEEE.

关键词： Requirements engineering

来源：评论

学校读者我要写书评

暂无评论

Voice Presentation Attacks Detection using Acoustic MLTP Features and BiLSTM 3

Voice Presentation Attacks Detection using Acoustic MLTP Fea...

引用

3rd International Conference on Communication, Computing and Digital Systems, C-CODE 2023

作者： Ibrar, Sundas Javed, Ali Ilyas, Hafsa Uet Taxila Department of Computer Science Punjab Pakistan Uet Taxila Department of Software Engineering Punjab Pakistan

ISBN: (纸本)9798350332391

Automatic speaker verification (ASV) systems have been used in numerous voice-controlled technologies including voice banking, automobiles, smartphones, smart homes, etc. The growth of voice-operated devices is increasing exponentially, and it is mandatory to secure the ASV systems against voice spoofing attacks such as cloned and synthetic voices. The research community has worked to develop spoofing countermeasures to improve the security of ASV systems. However, existing spoofing countermeasures have limitations to detect voice conversion, synthetic speech, and voice replay attacks. There still exists a need for the development of voice spoofing detectors to further strengthen the security of voice-controlled systems. Therefore, in this study, we designed a novel spoofing countermeasure for identifying physical access (PA) and logical access (LA) attacks. More precisely, we proposed a novel acoustic feature descriptor, namely, modified local ternary patterns (MLTP), and used them to train a BiLSTM classifier for voice spoofing detection. We assessed the performance of our approach using the standard ASVspoof-2019 dataset and attained an equal error rate (EER) of 1.87% and 2.66% for the PA and LA attacks, respectively. Experimental results obtained from a detailed assessment signify that our approach has the capability to accurately classify the bonafide and spoofed audio samples over contemporary methods. © 2023 IEEE.

关键词： Automation

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：