检索结果-内蒙古大学图书馆

13th International Workshop on Semantic Evaluation, SemEval 2019, co-located with the 17th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2019

作者： Fatyanosa, Tirana Noor Siagian, Al Hafiz Akbar Maulana Aritsugi, Masayoshi Computer Science and Electrical Engineering Graduate School of Science and Technology Kumamoto University Japan Big Data Science and Technology Faculty of Advanced Science and Technology Kumamoto University Japan Indonesian Institute of Sciences Indonesia

ISBN: (纸本)9781950737062

This paper describes the participation of DBMS-KU team in the SemEval 2019 Task 9, that is, suggestion mining from online reviews and forums. To deal with this task, we explore several machine learning approaches, i.e., Random Forest (RF), Logistic Regression (LR), Multinomial Naive Bayes (MNB), Linear Support Vector Classification (LSVC), Sublinear Support Vector Classification (SSVC), Convolutional Neural Network (CNN), and Variable Length Chromosome Genetic Algorithm-Naive Bayes (VLCGA-NB). Our system obtains reasonable results of F1-Score 0.47 and 0.37 on the evaluation data in Subtask A and Subtask B, respectively. In particular, our obtained results outperform the baseline in Subtask A. Interestingly, the results seem to show that our system could perform well in classifying Non-suggestion class. © 2019 Association for Computational Linguistics

关键词： Database systems

来源：评论

学校读者我要写书评

暂无评论

Self-paced resistance learning against overfitting on noisy labels

arXiv

引用

arXiv 2021年

作者： Shi, Xiaoshuang Guo, Zhenhua Li, Kang Liang, Yun Zhu, Xiaofeng School of Computer Science and Technology University of Electronic Science and Technology of China Sichuan Chengdu China Tsinghua Shenzhen International Graduate School Tsinghua University Guangdong Shenzhen China West China Medical Center Sichuan University Sichuan Chengdu China J. Crayton Pruitt Family Department of Biomedical Engineering University of Florida GainesvilleFL United States

Noisy labels composed of correct and corrupted ones are pervasive in practice. They might significantly deteriorate the performance of convolutional neural networks (CNNs), because CNNs are easily overfitted on corrupted labels. To address this issue, inspired by an observation, deep neural networks might first memorize the probably correct-label data and then corrupt-label samples, we propose a novel yet simple self-paced resistance framework to resist corrupted labels, without using any clean validation data. The proposed framework first utilizes the memorization effect of CNNs to learn a curriculum, which contains confident samples and provides meaningful supervision for other training samples. Then it adopts selected confident samples and a proposed resistance loss to update model parameters;the resistance loss tends to smooth model parameters’ update or attain equivalent prediction over each class, thereby resisting model overfitting on corrupted labels. Finally, we unify these two modules into a single loss function and optimize it in an alternative learning. Extensive experiments demonstrate the significantly superior performance of the proposed framework over recent state-of-the-art methods on noisy-label data. Source codes of the proposed method are available on https://***/xsshi2015/Self-paced-Resistance-Learning. Copyright © 2021, The Authors. All rights reserved.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Spectrograms Fusion-based End-to-end Robust Automatic Speech Recognition

Spectrograms Fusion-based End-to-end Robust Automatic Speech...

引用

Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)

作者： Hao Shi Longbiao Wang Sheng Li Cunhang Fan Jianwu Dang Tatsuya Kawahara Graduate School of Informatics Kyoto University Kyoto Japan Tianjin Key Laboratory of Cognitive Computing and Application College of Intelligence and Computing Tianjin University Tianjin China National Institute of Information and Communications Technology (NICT) Kyoto Japan Anhui Province Key Laboratory of Multimodal Cognitive Computation School of Computer Science and Technology Anhui University Japan Advanced Institute of Science and Technology Ishikawa Japan

ISBN: (纸本)9781665441629

To improve the robustness of automatic speech recognition (ASR), speech enhancement (SE) is often used as a front-end noise-removal process. Although there is complemen-tarity between the mapping-based and the mask-based SE system, one of the SE systems has been conventionally used as the front-end of ASR. We propose a spectrogram fusion (SF)-based end-to-end (E2E) robust ASR system, in which the mapping-based and masking-based SE are used as the front-end simultaneously. We adopt SF to combine the advantages of mapping-based and masking-based SE systems. SF and ASR modules are connected in an E2E manner, and joint training is conducted to finetune the front-end and the back-end. We compared the performance of different front-ends after joint training. From the experiments using Aishell and PNL 100 Nonspeech Sounds datasets, we found that the fusion of two SEs are beneficial for ASR, especially under low signal-to-noise ratio, where a relative improvement of more than 7% is achieved.

关键词： Training Information processing Speech enhancement Robustness Spectrogram Automatic speech recognition Signal to noise ratio

来源：评论

学校读者我要写书评

暂无评论

Experimental Analyses of an Efficient Aggregated Robot Processing with Cache-Control for Multi-Robot System

Experimental Analyses of an Efficient Aggregated Robot Proce...

引用

International Conference on Control, Automation and Systems ( ICCAS)

作者： Abdul Jalil Jun Kobayashi STMIK Handayani Makassar Indonesia Graduate School of Computer Science and Systems Engineering Kyushu Institute of Technology Iizuka Japan Faculty of Computer Science and Systems Engineering Kyushu Institute of Technology Iizuka Japan

ISBN: (数字)9788993215205

ISBN: (纸本)9781728185620

This study aims to develop a network architecture for low-cost multi-robot systems, in which robots operate with limited hardware capability and on-board computational power. Low-speed data processing and high latency time in communication are unavoidable issues in such a multi-robot system. To address those problems, this study proposes Aggregated Robot Processing - Robot Operating System (ARP-ROS) as a network architecture for low-cost multi-robot systems. This network architecture arranges a computing environment dedicated to data processing in a multi-robot system. In the proposed network architecture, most of sensor data obtained by each robot are transmitted to the computing environment via the communication network and processed there, and then control commands are sent back to the robots. For the data exchange in the network architecture, ROS2 node communication is employed because it uses Data Distribution Service (DDS) suitable for real-time distributed embedded systems. In addition, Cache-Control Algorithm (CCA) is introduced to the communication network for its performance improvement. This paper presents experimental analyses of the ARP-ROS with CCA in terms of latency time and data process failures.

关键词： Robots Network architecture Robot sensing systems Multi-robot systems computer architecture Data processing Distributed databases

来源：评论

学校读者我要写书评

暂无评论

Anomalous Shiba states in topological iron-based superconductors

arXiv

引用

arXiv 2022年

作者： Ghazaryan, Areg Kirmani, Ammar Fernandes, Rafael M. Ghaemi, Pouyan Institute of Science and Technology Austria Am Campus 1 Klosterneuberg3400 Austria Physics Department City College of the City University of New York NY10031 United States School of Physics and Astronomy University of Minnesota MinneapolisMN55455 United States Physics Program Graduate Center of City University of New York NY10031 United States

We demonstrate the formation of robust zero energy modes close to magnetic impurities in the iron-based superconductor FeSe1−xTex. We find that the Zeeman field generated by the impurity favors a spin-triplet inter-orbital pairing as opposed to the spin-singlet intra-orbital pairing prevalent in the bulk. The preferred spin-triplet pairing preserves time-reversal symmetry and is topological, as robust, topologically-protected zero modes emerge at the boundary between regions with different pairing states. Moreover, the zero modes form Kramers doublets that are insensitive to the direction of the spin polarization or to the separation between impurities. We argue that our theoretical results are consistent with recent experimental measurements on FeSe1−xTex. © 2022, CC BY.

关键词： Spin polarization

来源：评论

学校读者我要写书评

暂无评论

Privacy Preserving Blockchain Technique to Achieve Secure and Reliable Sharing of IoT Data

引用

computers, Materials & Continua 2020年第10期65卷 87-107页

作者： Bao Le Nguyen E.Laxmi Lydia Mohamed Elhoseny Irina V.Pustokhina Denis A.Pustokhin Mahmoud Mohamed Selim Gia Nhu Nguyen K.Shankar Faculty of Information Technology Duy Tan UniversityDa Nang550000Vietnam Computer Science and Engineering Vignan’s Institute of Information TechnologyVisakhapatnam530049India Faculty of Computers and Information Mansoura UniversityMansoura35516Egypt Department of Entrepreneurship and Logistics Plekhanov Russian University of EconomicsMoscow115093Russia Department of Logistics State University of ManagementMoscow109542Russia Department of Mathematics College of Science&Humanities in AlaflajPrince Sattam bin Abdulaziz UniversityAlaflaj16278Saudi Arabia Graduate School Duy Tan UniversityDa Nang550000Vietnam Department of Computer Applications Alagappa UniversityKaraikudi630003India

In present digital era,an exponential increase in Internet of Things(IoT)devices poses several design issues for business concerning security and *** studies indicate that the blockchain technology is found to be a significant solution to resolve the challenges of data security exist in *** this view,this paper presents a new privacy-preserving Secure Ant Colony optimization with Multi Kernel Support Vector Machine(ACOMKSVM)with Elliptical Curve cryptosystem(ECC)for secure and reliable IoT data *** program uses blockchain to ensure protection and integrity of some data while it has the technology to create secure ACOMKSVM training algorithms in partial views of IoT data,collected from various data ***,ECC is used to create effective and accurate privacy that protects ACOMKSVM secure learning *** this study,the authors deployed blockchain technique to create a secure and reliable data exchange platform across multiple data providers,where IoT data is encrypted and recorded in a distributed *** security analysis showed that the specific data ensures confidentiality of critical data from each data provider and protects the parameters of the ACOMKSVM model for data *** examine the performance of the proposed method,it is tested against two benchmark dataset such as Breast Cancer Wisconsin Data Set(BCWD)and Heart Disease Data Set(HDD)from UCI AI *** simulation outcome indicated that the ACOMKSVM model has outperformed all the compared methods under several aspects.

关键词： Blockchain optimization elliptical curve cryptosystem security ant colony optimization multi kernel support vector machine

来源：评论

学校读者我要写书评

暂无评论

PTeacher: A computer-Aided personalized pronunciation training system with exaggerated audio-visual corrective feedback

arXiv

引用

arXiv 2021年

作者： Bu, Yaohua Zhou, Hang Xu, Kaiyuan Yang, Zhihan Shi, Yuanchun Ma, Tianyi Jia, Jia Shi, Dachuan Li, Kun Lu, Xiaobo Li, Weijun Chen, Shengqi Wu, Haozhe Wu, Zhiyong Liu, Ziwei Academy of Arts & Design Tsinghua University The Chinese University of Hong Kong Department of Computer Science and Technology Tsinghua University SpeechX Limited School of Information Science and Technology Northeast Normal University Tsinghua-CUHK Joint Research Center for Media Sciences Technologies and Systems Shenzhen International Graduate School Tsinghua University S-Lab Nanyang Technological University

Second language (L2) English learners often find it difficult to improve their pronunciations due to the lack of expressive and personalized corrective feedback. In this paper, we present Pronunciation Teacher (PTeacher), a computer-Aided Pronunciation Training (CAPT) system that provides personalized exaggerated audio-visual corrective feedback for mispronunciations. Though the effectiveness of exaggerated feedback has been demonstrated, it is still unclear how to define the appropriate degrees of exaggeration when interacting with individual learners. To fill in this gap, we interview 100 L2 English learners and 22 professional native teachers to understand their needs and experiences. Three critical metrics are proposed for both learners and teachers to identify the best exaggeration levels in both audio and visual modalities. © 2021, CC BY.

关键词： Personnel training

来源：评论

学校读者我要写书评

暂无评论

The effect of AMC for each medium on QoE of video and audio transmission over wireless LANs 9

The effect of AMC for each medium on QoE of video and audio ...

引用

2019 9th International Workshop on computer science and Engineering, WCSE 2019 SPRING

作者： Nunome, Toshiro Murase, Naruhiro Department of Computer Science Graduate School of Engineering Nagoya Institute of Technology Nagoya466-8555 Japan Department of Computer Science Faculty of Engineering Nagoya Institute of Technology Nagoya466-8555 Japan

ISBN: (纸本)9789811414558

In this paper, we assess QoE of video and audio transmission over wireless LANs. This study tries to employ different AMC mechanis m for each medium. We consider four AMC methods: ARF, AARF, AARF-CD, and CARA. Through computer simulation and subjective experiment, we notice that the appropriate AMC for the required quality can enhance QoE. © Proceedings of 2019 the 9th International Workshop on computer science and Engineering, WCSE 2019 SPRING. All rights reserved.

关键词： Wireless local area networks (WLAN)

来源：评论

学校读者我要写书评

暂无评论

Effects of Built Environment on Bus Trip Rates under Rail Transit Competition

SSRN

引用

SSRN 2021年

作者： Wang, Shuxian Zhao, Liyuan Zhang, Ming Chen, Ruoyu Liang, Sicheng School of Architecture and Urban Planning HuaZhong University of Science and Technology Wuhan China Community and Regional Planning Program School of Architecture The University of Texas at Austin AustinTX78712 United States School of Urban Planning and Design Peking University Shenzhen Graduate School Shenzhen China Wuhan Planning and Design Institute Wuhan China

It's important to explore ways to integrate bus and rail transit by combining both the effects of the built environment and their competition. Using Wuhan city in China as a case study, this study develops gradient boosting decision trees(GBDTs) to examine the threshold range of built environment impacts on bus transit trip rates and to further explore the competitive impacts of rail transit on bus transit within these threshold ranges. The results show that when the number of service point-of-interests(POIs) within a 500 m radius of bus stop reaches 350 and the distance to the destination is 5 km, the model split of bus transit is the highest. The competition impact shows that rail transit promotes bus transit trip rates in urban fringe areas, but restricts in high-density central areas and the regions dominated by short distance trips. These findings provide technical support for optimizing spatial layouts through transit-oriented development. © 2021, The Authors. All rights reserved.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

Effects of Built Environment on Bus Trip Rates under Rail Transit Competition

SSRN

引用

SSRN 2021年

作者： Wang, Shuxian Zhao, Liyuan Zhang, Ming Chen, Ruoyu Liang, Sicheng School of Architecture and Urban Planning HuaZhong University of Science and Technology Wuhan China Community and Regional Planning Program School of Architecture University of Texas at Austin AustinTX78712 United States School of Urban Planning and Design Peking University Shenzhen Graduate School Shenzhen China Wuhan Planning and Design Institute Wuhan China

It’s important to explore ways to integrate bus and rail transit by combining both the effects of the built environment and their competition. Using Wuhan city in China as a case study, this study develops gradient boosting decision trees(GBDTs) to examine the threshold range of built environment impacts on bus transit trip rates and to further explore the competitive impacts of rail transit on bus transit within these threshold ranges. The results show that when the number of service point-of-interests(POIs) within a 500 m radius of bus stop reaches 350 and the distance to the destination is 5 km, the model split of bus transit is the highest. The competition impact shows that rail transit promotes bus transit trip rates in urban fringe areas, but restricts in high-density central areas and the regions dominated by short distance trips. These findings provide technical support for optimizing spatial layouts through transit-oriented development. © 2021, The Authors. All rights reserved.

关键词： Decision trees

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：