检索结果-内蒙古大学图书馆

Cyclic Transfer Learning for Mandarin-English Code-Switching Speech Recognition

IEEE Signal Processing Letters 2023年 30卷 1387-1391页

作者： Nga, Cao Hong Vu, Duc-Quang Luong, Huong Hoang Huang, Chien-Lin Wang, Jia-Ching National Central University Department of Computer Science and Information Engineering Taoyuan320 Taiwan Thai Nguyen University of Education Department of Computer Science and Information System Thai Nguyen7905218 Viet Nam Fpt University Department of Information Technology Can Tho900000 Viet Nam National Cheng Kung University Department of Computer Science and Information Engineering Tainan701 Taiwan

Transfer learning is a common method to improve the performance of the model on a target task via pre-training the model on pretext tasks. Different from the methods using monolingual corpora for pre-training, in this study, we propose a Cyclic Transfer Learning method (CTL) that utilizes both code-switching (CS) and monolingual speech resources as the pretext tasks. Moreover, the model in our approach is always alternately learned among these tasks. This helps our model can improve its performance via maintaining CS features during transferring knowledge. The experiment results on the standard SEAME Mandarin-English CS corpus have shown that our proposed CTL approach achieves the best performance with Mixed Error Rate (MER) of 16.3% on testman, 24.1% on testsge. In comparison to the baseline model that was pre-trained with monolingual data, our CTL method achieves 11.4% and 8.7% relative MER reduction on the testman and testsge sets, respectively. Besides, the CTL approach also outperforms compared to other state-of-the-art methods. © 1994-2012 IEEE.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Ga/GaSb nanostructures:Solution-phase growth for highperformance infrared photodetection

引用

Nano Research 2023年第2期16卷 3304-3311页

作者： Huanran Li Su You Yongqiang Yu Lin Ma Li Zhang Qing Yang Hefei National Research Center of Physical Sciences at the Microscale Department of ChemistryLaboratory of Nanomaterials for Energy Conversion(LNEC)University of Science and Technology of China(USTC)Hefei 230026China Micro Electromechanical System Research Center of Engineering and Technology of Anhui Province School of Electrical Science and Applied Physics Hefei University of Technology(HFUT)Hefei 230009China Nano Science and Technology Institute University of Science and Technology of China(USTC)Suzhou 215123China

Gallium antimonide(GaSb)-based nanostructures have been reported via various vapor-phase synthetic routes while there is not a report on the growth of GaSb nanostructures via a complete one-step solution-phase synthetic *** we report the design and synthesis of tadpole-like Ga/GaSb nanostructures by a one-step solution-phase synthetic route typically from the precursors of commercial triphenyl antimony(Sb(Ph)_(3))and trimethylaminogallium(Ga(NMe_(2))_(3))at 260°C in *** GaSb nanocrystals are grown based on a solution–liquid–solid(SLS)mechanism with zinc blende phase,and their size and shape can be controlled in the procedures via manipulating the reaction ***,the tadpole-like Ga/GaSb nanostructures can be applied for the fabrication of a GaSb/Si nanostructured heterojunction-like photodetector over silicon wafer,which demonstrates excellent photoresponse and detection performances from wavelength of 405 to 1,064 nm with high photoresponding ***,the photodetector exhibits a high responsivity of 18.9 A·W^(−1),a superior detectivity of 1.1×10^(13)Jones,and an ultrafast response speed of 44 *** present work provides a new strategy to group III–V antimonide-based semiconducting nanostructures that are capable for the fabrication of photodetector with broadband,high-detectivity,and high-speed photodetecting performances.

关键词： Ga/GaSb nanostructure metal-semiconductor heterojunction narrow bandgap semiconductor solution-liquid-solid(SLS)growth model GaSb/Si heterojunction photodetector hybrid nanostructured photodetector infrared photodetection

来源：评论

学校读者我要写书评

暂无评论

Pt thin-film resistance thermo detectors with stable interfaces for potential integration in SiC hightemperature pressure sensors

引用

Microsystems & Nanoengineering 2024年第5期10卷 231-242页

作者： Ziyan Fang Xiaoyu Wu Hu Zhao Xudong Fang Chen Wu Dong Zhang Zhongkai Zhang Bian Tian Libo Zhao Tiefu Li Prateek Verma Ryutaro Maeda Zhuangde Jiang State Key Laboratory for Manufacturing Systems Engineering International Joint Laboratory for Micro/Nano Manufacturing and Measurement TechnologiesXi'an Jiaotong UniversityXi'an 710049China School of Mechanical Engineering Xi'an Jiaotong UniversityXi'an 710049China Shandong Laboratory of Yantai Advanced Materials and Green Manufacturing at Yantai Yantai 264000China Xi'an Jiaotong University(Yantai)Research Institute for Intelligent Sensing Technology and System Xi'anChina School of Instrument Science and Technology Xi'an Jiaotong UniversityXi'an 710049China School of Integrated Circuits Tsinghua UniversityBeijing 100084China Department of Chemical Engineering University of ArkansasFayettevilleAR72701USA

Due to the excellent mechanical,chemical,and electrical properties of third-generation semiconductor silicon carbide(SiC),pressure sensors utilizing this material might be able to operate in extreme environments with temperatures exceeding 300℃.However,the significant output drift at elevated temperatures challenges the precision and stability of ***-time in situ temperature monitoring of the pressure sensor chip is highly important for the accurate compensation of the pressure *** this study,we fabricate platinum(Pt)thin-film resistance temperature detectors(RTDs)on a SiC substrate by incorporating aluminum oxide(Al_(2)O_(3))as the transition layer and utilizing aluminum nitride(AIN)grooves for alignment through microfabrication *** composite layers strongly adhere to the substrate at temperatures reaching 950℃,and the interface of the Al2O3/Pt bilayer remains stable at elevated temperatures of approximately 950 *** stability contributes to the excellent high-temperature electrical performance of the Pt RTD,enabling it to endure temperatures exceeding 850℃ with good *** characteristics establish a basis for the future integration of Pt RTD in SiC pressure ***,tests and analyses are conducted on the interfacial diffusion,surface morphological,microstructural,and electrical properties of the Pt flms at various annealing *** can be inferred that the tensile stress and self-diffusion of Pt films lead to the formation of hillocks,ultimately reducing the electrical performance of the Pt thin-flm *** increase the upper temperature threshold,steps should be taken to prevent the agglomeration of Pt films.

关键词： resistance carbide film

来源：评论

学校读者我要写书评

暂无评论

TRACING THE MOTION IN HIGH-SPEED VIDEO FRAMES USING DEEP LEARNING

TRACING THE MOTION IN HIGH-SPEED VIDEO FRAMES USING DEEP LEA...

引用

2023 IEEE International Conference on Research Methodologies in Knowledge Management, Artificial Intelligence and Telecommunication engineering, RMKMATE 2023

作者： Gayathri, R. Harsha, B.K. Sharmiladevi, D. Sajitha, L.P. Rao, Vuda Sreenivasa Nirmala, G. Department of ECE Rajalakshmi Engineering College Tamilnadu Chennai India Department of ECE CMR Institute of Technology Karnataka Bengaluru India Department of Electronics and Communication Engineering Sona College of Technology Tamilnadu Salem India Department of Computer Science and Business System R. M. K Engineering College Tamilnadu Kavaraipettai India Department of CSE Koneru Lakshmaiah Education Foundation AP Vaddeswaram India Department of Computer Science and Engineering R.M.D Engineering College Chennai India

ISBN: (纸本)9798350305708

One of the most important jobs of a self-driving vehicle's visual system is to identify and categorize obstacles along the route. Cars need to be detected, categorized, localized, and linked across frames if they are to be tracked. With the low cost and high utility of visual information, vision-based technologies have gained popularity for this purpose. In this paper, we develop a deep learning-based motion tracking from high-speed video frames. The model is developed using deep convolutional neural network that makes it to be applicable for classification of objects in the video frames. The deep convolutional neural network classifies the obstacles present in the pathway and offers a better classification pattern than the conventional models. © 2023 IEEE.

关键词： Application Convolutional Neural Network Deep Learning High-Speed Video Motion Vehicle Video Frames

来源：评论

学校读者我要写书评

暂无评论

Timescales and contribution of heating and helicity effect in helicity-dependent all-optical switching

引用

Rare Metals 2023年第1期42卷 234-243页

作者： Guan-Qi Li Xiang-Yu Zheng Jun-Lin Wang Xian-Yang Lu Jing Wu Jian-Wang Cai Hao Meng Bo Liu Thomas A.Ostler Yong-Bing Xu School of Integrated Circuits Guangdong University of TechnologyGuangzhou510006China York-Nanjing International Joint Center in Spintronics School of Electronic Science and EngineeringNanjing UniversityNanjing210093China Spintronics and Nanodevice Laboratory Department of Electronics EngineeringUniversity of YorkYorkYO105DDUK Institute of Physics Chinese Academy of SciencesBeijing100190China Key Laboratory of Spintronics Materials Devices and System of Zhejiang ProvinceHangzhou311305China College of Business Technology and EngineeringSheffield Hallam UniversitySheffieldS11WBUK

The heating and helicity effects induced by circularly polarized laser excitation are entangled in the helicity-dependent all-optical switching(HD-AOS),which hinders understanding the magnetization dynamics ***,applying a dual-pump laser excitation,first with a linearly polarized(LP) laser pulse followed by a circularly polarized(CP) laser pulse,the timescales and contribution from heating and helicity effects in HD-AOS were identified with a Pt/Co/Pt *** the LP laser pulses preheat the sample to a nearly fully demagnetized state,the CP laser pulses with a power reduced by 80% switch the sample’s *** varying the time delay between the two pump pulses,the results show that the helicity effect,which gives rise to the deterministic helicity-induced switching,arises almost instantly within 200 fs close to the pulse width upon laser *** results reveal that the transient magnetization state upon which CP laser pulses impinge is the key factor for achieving HD-AOS,and importantly,the tunability between heating and helicity effects with the unique dualpump laser excitation approach will enable HD-AOS in a wide range of magnetic material systems having wideranging implications for potential ultrafast spintronics applications.

关键词： Helicity-dependent all-optical switching Ultrafast laser Ultrafast magnetization Magneto-optical interaction

来源：评论

学校读者我要写书评

暂无评论

Empirical Assessment and Detection of Suicide Related Posts in Twitter using Artificial Intelligence enabled Classification Logic 2

Empirical Assessment and Detection of Suicide Related Posts ...

引用

2nd IEEE International Conference on Advances in Computing, Communication and Applied Informatics, ACCAI 2023

作者： Ravishankar, T. Nadana Kumar, Ata. Kishore Venkatesh, J. Prabhu, M.Ramkumar Bhargavi, V. Sharmila Muthamilselvan, S. Srm Institute of Science and Technology Department of Data Science and Business System Kattankulathur India School of Engineering Department of ECE Tirupati India Chennai Institute of Technology Department of Cse Chennai India Peri Institute of Technology Department of ECE Chennai India R.M.K. Engineering College Department of Cse Chennai India Sri Sairam Engineering College Department of Artificial Intelligence & Data Science Chennai India

ISBN: (纸本)9798350315905

The identification of suicidal thoughts in online social networks is an expanding field of study fraught with major challenges. Recent studies have shown that the readily available data, dispersed over many online life phases, contains useful clues for accurately identifying persons with suicidal intentions. The primary challenge in preventing suicide is learning to recognize and respond appropriately to the sometimes-confusing risk factors and warning indications that may precipitate an attempt. Indicators useful for diagnosing people with suicide thoughts can be found in publicly available material shared over social media platforms, according to recent studies. Understanding and recognizing the myriad risk factors and warning symptoms that may precede a suicide attempt is the primary difficulty in this area of public health. In this research, we developed a benchmark for multi-class categorization using machine learning models. We used a majority classifier, a frequency-based technique, and two deep learning models as our models. Both deep learning models outperformed the majority and the word frequency classifier, with results that were very comparable. These classification results are on par with the state-of-the-art on similar problems and, in most cases, with human results. © 2023 IEEE.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Optimized Moderator Design and Analysis of a Pin-Type Supercritical Carbon Dioxide Reactor Based on Reactor Monte Carlo Code

Optimized Moderator Design and Analysis of a Pin-Type Superc...

引用

作者： Zhao, Xingyu Liu, Minyun Cui, Rongyi Huang, Shanfang Wang, Kan Lu, Chuan Department of Engineering Physics Tsinghua University Beijing100084 China Department of Engineering Physics Tsinghua University Beijing100084 China China Nuclear Power Engineering Co. Ltd. Beijing100840 China Science and Technology on Reactor System Design Technology Laboratory Nuclear Power Institute of China Chengdu610213 China

This study analyzed an yttrium hydride (YH2) moderated supercritical carbon dioxide cooled reactor loaded with pin-type, beryllium oxide diluted oxide fuel elements to reduce the critical enrichment. The impact of the YH2 on the coolant void reactivity was studied along with a moderator zoning scheme to flatten the radial power distribution. The YH2 was added as hexagonal moderating rods at the center of the fuel assemblies. The core was modeled using the continuous-energy Reactor Monte Carlo code (RMC) with the on-the-fly cross sections treatment. The results showed that the YH2 moderator increased the thermal fission and reduced the critical enrichment of the core with the same diluent volume fraction by more than 30%. The YH2 moderator significantly softened the neutron energy spectrum and reduced the neutron leakage upon core voiding, resulting in both a weaker positive spectral reactivity feedback and a weaker negative leakage reactivity feedback during core depressurization. For an UO2-loaded core, the YH2 gave a lower negative coolant void reactivity, while for a mixed oxide fuel (MOX)-loaded core with diluent volume fractions smaller than 35%, the spectral feedback was more important and the YH2 strongly reduced the positive coolant void reactivity to less than $1. Arranging the YH2 in the peripheral assemblies reduced the radial power peaking factor to 1.319. The study shows that the YH2 moderator can reduce the critical enrichment, make the core less sensitive to voiding, and can flatten the radial power distribution of a single-enrichment core through moderator zoning. [DOI: 10.1115/1.4056772] © 2023 American Society of Mechanical Engineers (ASME). All rights reserved.

关键词： Carbon dioxide

来源：评论

学校读者我要写书评

暂无评论

Development of a high-resolution near-surface meteorological forcing dataset for the Third Pole region

引用

Science China Earth Sciences 2025年第4期68卷 1274-1290页

作者： Yaozhi JIANG Wenjun TANG Kun YANG Jie HE Changkun SHAO Xu ZHOU Hui LU Yingying CHEN Xin LI Jiancheng SHI School of Resources and Environment University of Electronic Science and Technology of ChinaChengdu611731China Department of Earth System Science Ministry of Education Key Laboratory for Earth System ModelingInstitute for Global Change StudiesTsinghua UniversityBeijing100084China National Tibetan Plateau Data Center State Key Laboratory of Tibetan Plateau Earth SystemEnvironment and ResourcesInstitute of Tibetan Plateau ResearchChinese Academy of SciencesBeijing100101China State Key Laboratory of Hydroscience and Engineering Tsinghua UniversityBeijing100084China Tsinghua University(Department of Earth System Science)-Xi’an Institute of Surveying and Mapping Joint Research Center for Next-Generation Smart Mapping Beijing100084China National Space Science Center Chinese Academy of SciencesBeijing100190China

Recently,the Third Pole(TP)region has experienced rapid environmental *** data are essential for hydrometeorological and ecological applications but still have large uncertainties on the TP owing to the heterogeneous land surface,complex terrain,and sparse weather *** this study,a long-term(1979–2020)high-resolution(1/30°)meteorological forcing dataset for the TP(TPMFD)was developed,as a sister to the widely used China Meteorological Forcing Dataset(CMFD).The TPMFD comprises seven components necessary for driving land surface *** have previously contributed precipitation and downward shortwave radiation data for the TPMFD,and this study presents the development of five other components and focuses on validations for all ***,2-meter air temperature,2-meter specific humidity,10-meter wind speed,and surface air pressure were generated by combining the fifth-generation atmospheric reanalysis for European Center for Medium-Range Weather Forecasts(ERA5),a short-term high-resolution atmospheric simulation,and in situ observations,and the downward longwave radiation was calculated using semi-physical *** cross-validation and independent-validation demonstrated that most variables in the developed dataset outperformed those in widely used reanalysis datasets,including ERA5,ERA5-Land,and the Global Land Data Assimilation system(GLDAS).This dataset is expected to be beneficial for climate analyses and modeling applications of land-surface processes on the TP.

关键词： Meteorological data Third Pole High-resolution Data downscaling and merging

来源：评论

学校读者我要写书评

暂无评论

A Novel Method for Accurate & Real-time Food Classification: The Synergistic Integration of EfficientNetB7, CBAM, Transfer Learning, and Data Augmentation

arXiv

引用

arXiv 2024年

作者： Rokhva, Shayan Teimourpour, Babak Department of Information Technology Engineering Faculty of Industrial and System Engineering Tarbiat Modares University Tehran Iran

Integrating artificial intelligence into modern society is profoundly transformative, significantly enhancing productivity by streamlining various daily tasks. AI-driven recognition systems provide notable advantages in the food sector, including improved nutrient tracking, tackling food waste, and boosting food production and consumption efficiency. Accurate food classification is a crucial initial step in utilizing advanced AI models, as the effectiveness of this process directly influences the success of subsequent operations;therefore, achieving high accuracy at a reasonable speed is essential. Despite existing research efforts, a gap persists in improving performance while ensuring rapid processing times, prompting researchers to pursue cost-effective and precise models. This study addresses this gap by employing the state-of-the-art EfficientNetB7 architecture, enhanced through transfer learning, data augmentation, and the CBAM attention module. This methodology results in a robust model that surpasses previous studies in accuracy while maintaining rapid processing suitable for real-world applications. The Food11 dataset from Kaggle was utilized, comprising 16643 imbalanced images across 11 diverse classes with significant intra-category diversities and inter-category similarities. Furthermore, the proposed methodology, bolstered by various deep learning techniques, consistently achieves an impressive average accuracy of 96.40%. Notably, it can classify over 60 images within one second during inference on unseen data, demonstrating its ability to deliver high accuracy promptly. This underscores its potential for practical applications in accurate food classification and enhancing efficiency in subsequent processes. © 2024, CC BY.

关键词： Data accuracy

来源：评论

学校读者我要写书评

暂无评论

Data-driven Power Flow Method Based on Exact Linear Regression Equations

引用

Journal of Modern Power systems and Clean Energy 2022年第3期10卷 800-804页

作者： Yanbo Chen Chao Wu Junjian Qi State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources School of Electrical&Electronic EngineeringNorth China Electric Power UniversityBeijing 102206China Department of Electrical and Computer Engineering Stevens Institute of TechnologyHobokenNJ 07030USA

Power flow(PF)is one of the most important calculations in power *** widely-used PF methods are the Newton-Raphson PF(NRPF)method and the fast-decoupled PF(FDPF)*** smart grids,power generations and loads become intermittent and much more uncertain,and the topology also changes more frequently,which may result in significant state shifts and further make NRPF or FDPF difficult to *** address this problem,we propose a data-driven PF(DDPF)method based on historical/simulated data that includes an offline learning stage and an online computing *** the offline learning stage,a learning model is constructed based on the proposed exact linear regression equations,and then the proposed learning model is solved by the ridge regression(RR)method to suppress the effect of data *** online computing stage,the nonlinear iterative calculation is not *** results demonstrate that the proposed DDPF method has no convergence problem and has much higher calculation efficiency than NRPF or FDPF while ensuring similar calculation accuracy.

关键词： Data driven exact linear regression equation Fast-decoupled power flow Newton-Raphson method

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：