检索结果-内蒙古大学图书馆

IEEE Conference on Global Communications (GLOBECOM)

作者： Minyang Li Xiangyun Tang Tao Zhang Guangyuan Liu Yu Weng Key Laboratory of Ethnic Language Intelligent Analysis and Security Governance Minzu University of China China School of Software Engineering Beijing Jiaotong University China College of Computing and Data Science Nanyang Technological University Singapore

ISBN: (数字)9798350351255

ISBN: (纸本)9798350351262

Federated Learning (FL) is a distributed machine learning framework that enhances privacy by enabling multiple participants to train a global model without sharing their raw data. However, FL still faces the threat posed by data and Byzantine attacks (e.g., data poisoning and model poisoning attacks) from malicious clients aimed to decrease the FL global training accuracy. Previous studies have proposed solutions to improve the robustness of FL systems. However, these robust approaches have not effectively defended against Byzantine attacks from malicious clients in non-Independent and Identically Distributed (non-IID) environments, decreasing overall training accuracy. In this work to address this issues, we propose a new defence framework that aims to enhance model accuracy against Byzantine attacks from malicious clients in non-IID environments. In our approach, the central server performs the Inverse Deep Learning Gradient attack using the gradient data submitted by users in each round, obtaining an inversed artificial gradient. We then assess the squared L2 norm difference between this inversed gradient and the actual gradients to detect malicious clients. Simultaneously, we assign weights to each client involved in the aggregation based on differences in the squared L2 norm, aiming to minimize the impact of malicious actors on the overall model. We have experimentally evaluated our method and demonstrated its ability to maintain training accuracy under attack in non-IID environments using standard datasets.

关键词： Training Accuracy Federated learning Distributed databases Robustness Data models Servers Global communication Standards Faces

来源：评论

学校读者我要写书评

暂无评论

An AHP-based Computer Evaluation Model for College Students English Pronunciation Quality 21

An AHP-based Computer Evaluation Model for College Students ...

引用

5th International Conference on Electronic Information Technology and Computer engineering, EITCE 2021

作者： Long, Yan Li, Xinguang Chen, Shuai Zhang, Shenbin Ma, Shanxian Li, Ruisi Laboratory of Language Engineering and Computing School of English Education Guangdong University of Foreign Studies Guangzhou China Laboratory of Language Engineering and Computing School of Information Science and Technology Guangdong University of Foreign Studies Guangzhou China School of Information Science and Technology Guangdong University of Foreign Studies Guangzhou China College of Computer Science South-Central University for Nationalities Wuhan China

ISBN: (纸本)9781450384322

Different groups of English learners have different levels of spoken English, and the criteria used to evaluate their pronunciation should also be different. This paper introduces a method for determining the weights of computerized ratings of English pronunciation quality for university students using the Analytic Hierarchy Process (AHP). The indicators used for scoring are divided into two parts: firstly, basic pronunciation, including accuracy, intensity and speed;secondly, expression ability, including emotion, rhythm and intonation. The best weight for each indicator is calculated by the AHP method, and when the evaluation is carried out, the weighted sum of the scores of each indicator is calculated to obtain the final overall evaluation result. The paper concludes with an experiment to verify that the computer evaluation model of English pronunciation quality established in this way for university students is objective and valid. © 2021 ACM.

关键词： Analytic hierarchy process

来源：评论

学校读者我要写书评

暂无评论

The Application of Bi-LSTM in Computerized Adaptive Test Research

The Application of Bi-LSTM in Computerized Adaptive Test Res...

引用

2020 International Conference on Big Data and Social Sciences, ICBDSS 2020

作者： Chen, Jin Lin, Jianghao Guangdong University of Foreign Studies Laboratory for Language Engineering and Computing Guangzhou China

ISBN: (纸本)9781728197517

The study proposed an adaptive cognitive diagnostic model based on Bi-LSTM which regards the model as a sequence prediction from participants' responses to their knowledge state. With the merit of bidirectional long-sequence memory, Bi-LSTM is able to employ the whole sequence of response as the input of the adaptive cognitive diagnostic model. Comparing to traditional approaches with a maximum probability as the result of cognitive diagnosis, the model based on Bi-LSTM is superior in that it makes use of the whole response sequence to predict the result the model was verified through the actual data from 1029 British middle school students who participated in the English reading test of the Program for International Student Assessment (PISA). It is found that the accuracy rate and the average attribute accuracy rate of the model make significant breakthroughs. Meanwhile, the order of the test items has some influence on participants' responses, which may be considered as an impact factor in future test. © 2020 IEEE.

关键词： Software testing

来源：评论

学校读者我要写书评

暂无评论

Scored and Error-annotated Essay Dataset of Chinese EFL/ESL Learners 21

Scored and Error-annotated Essay Dataset of Chinese EFL/ESL ...

引用

Proceedings of the 2021 5th International Conference on Natural language Processing and Information Retrieval

作者： Kai Jin Wuying Liu School of Foreign Languages Qilu University of Technology China Laboratory of Language Engineering and Computing Guangdong University of Foreign Studies China

ISBN: (纸本)9781450387354

A certain scale of finely annotated essay dataset of EFL/ESL (English as a foreign language or the second language) learners is not only an important language resource for language research and teaching, but also contributing materials for language-related computing science. Unfortunately, this type of data open on the Internet are not only of small quantity but also of uneven quality, especially such data of Chinese learners. We collected 147 essays of Chinese EFL/ESL learners and had four teachers score them under the same criteria and one teacher annotate major errors, and have them scored in Pigai scoring system. We then structured the score file, error-annotated files, essay files together with context information, and built the Scored and Error-annotated Essay Dataset of Chinese EFL/ESL Learners (SeedCel) which is open on the Internet and will be incrementally updated. This paper explains how SeedCel is constructed, what the details of SeedCel are, and where SeedCel will be used.

关键词： Chinese leaners EFL/ESL Learners Error-annotated Essay Dataset Scored

来源：评论

学校读者我要写书评

暂无评论

Heart Sound Classification based on Fractional Fourier Transformation Entropy 4

Heart Sound Classification based on Fractional Fourier Trans...

引用

4th IEEE Global Conference on Life Sciences and Technologies, LifeTech 2022

作者： Tan, Yang Wang, Zhihua Qian, Kun Hu, Bin Zhao, Shiliang Schuller, Bjorn W. Yamamoto, Yoshiharu Beijing Institute of Technology School of Medical Technology China Educational Physiology Laboratory The University of Tokyo Japan School of Mechatronic Engineering China University of Mining and Technology China School of Physics and Electronic Engineering Sichuan Normal University China GLAM-Group on Language Audio & Music Department of Computing Imperial College London United Kingdom

ISBN: (纸本)9781665419048

Automatic classification of heart sounds has been studied for many years, because computer-aided auscultation of heart sounds can help doctors make a preliminary diagnosis. We propose a classification method for heart sounds that uses fractional Fourier transformation entropy (FRFE) as the features and a support vector machine (SVM) as the classification model. The process of the whole method is cutting of heart sounds, feature extraction, and classification. We compare FRFE of different signal orders, and finally evaluate fused features of multiple orders according to the better classification results. These fused features are used as the input of a SVM, a k-nearest neighbour (KNN), and a Naive bayes classifier (NBC) to compare the most suitable classifiers. Finally, we consider the fused features that reflect both the time and the frequency domain to achieve a better classification performance. © 2022 IEEE.

关键词： Entropy

来源：评论

学校读者我要写书评

暂无评论

Learning to Learn Transferable Generative Attack for Person Re-Identification

arXiv

引用

arXiv 2024年

作者： Bian, Yuan Liu, Min Wang, Xueping Ma, Yunfeng Wang, Yaonan College of Electrical and Information Engineering Hunan University National Engineering Research Center of Robot Visual Perception and Control Technology Hunan Changsha China College of Information Science and Engineering Hunan Normal University Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing Hunan Changsha China

Deep learning-based person re-identification (re-id) models are widely employed in surveillance systems and inevitably inherit the vulnerability of deep networks to adversarial attacks. Existing attacks merely consider cross-dataset and cross-model transferability, ignoring the cross-test capability to perturb models trained in different domains. To powerfully examine the robustness of real-world re-id models, the Meta Transferable Generative Attack (MTGA) method is proposed, which adopts meta-learning optimization to promote the generative attacker producing highly transferable adversarial examples by learning comprehensively simulated transfer-based cross-model&dataset&test black-box meta attack tasks. Specifically, cross-model&dataset black-box attack tasks are first mimicked by selecting different re-id models and datasets for meta-train and meta-test attack processes. As different models may focus on different feature regions, the Perturbation Random Erasing module is further devised to prevent the attacker from learning to only corrupt model-specific features. To boost the attacker learning to possess cross-test transferability, the Normalization Mix strategy is introduced to imitate diverse feature embedding spaces by mixing multi-domain statistics of target models. Extensive experiments show the superiority of MTGA, especially in cross-model&dataset and cross-model&dataset&test attacks, our MTGA outperforms the SOTA methods by 20.0% and 11.3% on mean mAP drop rate, respectively. The source codes are available at https://***/yuanbianGit/MTGA. Copyright © 2024, 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

Analyzing Relational Semantics of Clauses in Chinese Discourse Based on Feature Structure 9th

Analyzing Relational Semantics of Clauses in Chinese Discour...

引用

9th CCF International Conference on Natural language Processing and Chinese computing, NLPCC 2020

作者： Feng, Wenhe Huang, Xi Ren, Han Laboratory of Language Engineering and Computing Guangdong University of Foreign Studies Guangzhou510420 China

ISBN: (纸本)9783030604561

The discourse clause relational semantics is the semantic relation between discourse clause relevance structures. This paper proposes a method to represent the discourse clause relational semantics as a multi-dimensional feature structure. Compared with the simple classification mechanism of discourse relations, it can reveal the discourse semantic relations more deeply. Furthermore, we built Chinese discourse clause relational semantic feature corpus, and study the clause relational semantic feature recognition. We Transfer the clause relational semantic feature recognition into multiple binary classification problems, and extract relevant classification features for experiment. Experiments show that under the best classifier (SVM), the overall semantic feature recognition effect of F1 value reaches 70.14%;each classification feature contributes differently to the recognition of different clause relational semantic features, and the connectives contributes more to the recognition of all semantic features. By adding related semantic features as classification features, the interaction between different semantic features is studied. Experiments show that the influence of different semantic features is different. The addition of multiple semantic features has a more significant effect than a single semantic feature. © 2020, Springer Nature Switzerland AG.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A new type of Chinese speech synthesis technology and system research

A new type of Chinese speech synthesis technology and system...

引用

2022 International Conference on Electronic Information engineering, Big Data, and Computer Technology, EIBDCT 2022

作者： Li, Xinguang Liang, Chuhua Ma, Shanxian Liu, Congcong Chen, Shuai Li, Ruisi He, Haoxin Laboratory of Language Engineering and Computing Guangdong University of Foreign Studies Guangdong Guangzhou510006 China School of Information Science and Technology Guangdong University of Foreign Studies Guangdong Guangzhou510006 China College of Computer Science South-Central University for Nationalities Hubei Wuhan430074 China

ISBN: (纸本)9781510655133

In this paper, we propose a method of Chinese speech synthesis. In order to achieve the purpose of synthesizing fluent and natural speech, two processing methods, rule-based and statistics-based, are used in the developed system. Firstly, a module design method with the function of text language recognition is introduced in this paper. This module can classify and recognize the text of Chinese, English and special symbols, and deal with the recognition problem that the input text contains Chinese, English and special symbols. Secondly, the Chinese speech synthesis methods used in the system are explained. In the prosody control module, we use a prosody structure prediction model combining neural network and decision tree;when concatenating speech, we propose two smoothing method. One is smoothing algorithm based on spectrum analysis and the other is smoothing algorithm based on dual-threshold. Experiments prove that the synthesis effect of our system is great, and the synthesized speech is clear and natural. Due to the moderate amount of system data and high code efficiency, it is suitable for application systems in mobile Android. © 2022 SPIE.

关键词： Speech synthesis

来源：评论

学校读者我要写书评

暂无评论

Modality Unified Attack for Omni-Modality Person Re-Identification

arXiv

引用

arXiv 2025年

作者： Bian, Yuan Liu, Min Yi, Yunqi Wang, Xueping Ma, Yunfeng Wang, Yaonan the College of Electrical and Information Engineering Hu nan University National Engineering Research Center of Robot Vi sual Perception and Control Technology Hunan Changsha China the College of Information Science and Engineering Hunan Normal University Hunan Provincial Key Laboratory of Intel ligent Computing and Language Information Processing Hunan Changsha China

Deep learning based person re-identification (reid) models have been widely employed in surveillance systems. Recent studies have demonstrated that black-box single-modality and cross-modality re-id models are vulnerable to adversarial examples (AEs), leaving the robustness of multi-modality reid models unexplored. Due to the lack of knowledge about the specific type of model deployed in the target black-box surveillance system, we aim to generate modality unified AEs for omni-modality (single-, cross- and multi-modality) re-id models. Specifically, we propose a novel Modality Unified Attack method to train modality-specific adversarial generators to generate AEs that effectively attack different omni-modality models. A multi-modality model is adopted as the surrogate model, wherein the features of each modality are perturbed by metric disruption loss before fusion. To collapse the common features of omni-modality models, Cross Modality Simulated Disruption approach is introduced to mimic the cross-modality feature embeddings by intentionally feeding images to non-corresponding modality-specific subnetworks of the surrogate model. Moreover, Multi Modality Collaborative Disruption strategy is devised to facilitate the attacker to comprehensively corrupt the informative content of person images by leveraging a multi modality feature collaborative metric disruption loss. Extensive experiments show that our MUA method can effectively attack the omni-modality re-id models, achieving 55.9%, 24.4%, 49.0% and 62.7% mean mAP Drop Rate, respectively. Copyright © 2025, The Authors. All rights reserved.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Deep progressive reinforcement learning-based flexible resource scheduling framework for IRS and UAV-assisted MEC system

arXiv

引用

arXiv 2024年

作者： Dong, Li Jiang, Feibo Wang, Minjie Peng, Yubo Li, Xiaolong School of Computer Science Hunan University of Technology and Business Changsha China Xiangjiang Laboratory Changsha China Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing Hunan Normal University Changsha China College of Information Science and Engineering Hunan Normal University Changsha China

The intelligent reflection surface (IRS) and unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) system is widely used in temporary and emergency scenarios. Our goal is to minimize the energy consumption of the MEC system by jointly optimizing UAV locations, IRS phase shift, task offloading, and resource allocation with a variable number of UAVs. To this end, we propose a Flexible REsource Scheduling (FRES) framework by employing a novel deep progressive reinforcement learning which includes the following innovations: Firstly, a novel multi-task agent is presented to deal with the mixed integer nonlinear programming (MINLP) problem. The multi-task agent has two output heads designed for different tasks, in which a classified head is employed to make offloading decisions with integer variables while a fitting head is applied to solve resource allocation with continuous variables. Secondly, a progressive scheduler is introduced to adapt the agent to the varying number of UAVs by progressively adjusting a part of neurons in the agent. This structure can naturally accumulate experiences and be immune to catastrophic forgetting. Finally, a light taboo search (LTS) is introduced to enhance the global search of the FRES. The numerical results demonstrate the superiority of the FRES framework which can make real-time and optimal resource scheduling even in dynamic MEC systems. Copyright © 2024, The Authors. All rights reserved.

关键词： Unmanned aerial vehicles (UAV)

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：