检索结果-南通市图书馆

SWG: an architecture for sparse weight gradient computation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第2期67卷 302-321页

作者： weiwei WU Fengbin TU xiangyu LI Shaojun wei Shouyi YIN School of Integrated Circuits Tsinghua University Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology

On-device training for deep neural networks(DNN) has become a trend due to various user preferences and scenarios. The DNN training process consists of three phases, feedforward(FF), backpropagation(BP), and weight gradient(WG) update. WG takes about one-third of the computation in the whole training process. Current training accelerators usually ignore the special computation property of WG and process it in a way similar to FF/BP. Besides, the extensive data sparsity existing in WG, which brings opportunities to save computation, is not well explored. Nevertheless, exploiting the optimization opportunities would meet three underutilization problems, which are caused by(1) the mismatch between WG data dimensions and hardware parallelism,(2) the full sparsity, i.e., the sparsity of feature map(Fmap),error map(Emap), and gradient, and(3) the workload imbalance resulting from irregular sparsity. In this paper, we propose a specific architecture for sparse weight gradient(SWG) computation. The architecture is designed based on hierarchical unrolling and sparsity-aware(HUSA) dataflow to exploit the optimization opportunities of the special computation property and full data sparsity. In HUSA dataflow, the data dimensions are unrolled hierarchically on the hardware architecture. A valid-data trace(VDT) mechanism is embedded in the dataflow to avoid the underutilization caused by the two-sided input sparsity. The gradient is unrolled in PE to alleviate the underutilization induced by output sparsity while maintaining the data reuse opportunities. Besides, we design an intra-and inter-column balancer(IIBLC) to dynamically tackle the workload imbalance problem resulting from the irregular sparsity. Experimental results show that with HUSA dataflow exploiting the full sparsity, SWG achieves a speedup of 12.23× over state-of-the-art gradient computation architecture, Train Ware. SWG helps to improve the energy efficiency of the state-of-the-art training accelerator LNPU from 7.56 to 10.58 TOPS/W.

关键词： CNN training gradient computation sparsity architecture

Research Progress and Challenges on Persistent Organic Pollutants in Lakes

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Journal of Earth Science 2024年第2期35卷 729-736页

作者： wei Guo Xiaoyu Ji Zhengfei Yu Hongchen Jiang xiangyu Guan School of Ocean Sciences China University of GeosciencesBeijing 100083China State Key Lab of Biogeology and Environmental Geology China University of GeosciencesWuhan 430074China

Lakes are the main reservoirs of persistent organic pollutants(POPs)from land,atmo sphere and rivers.POPs in lakes undergo complex exchange,transformation,and degradation between water-air-sediment-biota interfaces,which are constrained and regulated by various physical,chemical and biological factors.POPs can affect ecological conditions,chemical properties of water and sedi ments,and biodiversity of the lake system.Therefore,it is important to study the sources,migration,transformation,environmental behavior and ecological impacts of POPs in lake ecosystems.This re view summarizes research progress on detection technologies,diversity and origins,historical records,migration and transformation,distribution patterns,degradation and toxic effects of POPs in lakes.Fi nally,future directions related to POPs in lakes were summarized.

关键词： persistent organic pollutants(POPs) lakes transformation degradation microbes environment geolgogy

维普期刊数据库

Direct Analysis of Whole Blood by a Disposable Monolithic Column Mass Spectrometry Analysis Kit

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Chemistry 2024年第10期42卷 1073-1078页

作者： Wenxin Wang Jin Han Yutong Zheng Haijie Wang Chengbiao Zhu Yongyi Li Yuqi Feng Zhenwei wei xiangyu Wang College of Chemistry and Molecular Science Wuhan UniversityWuhanHubei 430072China School of Public Health Wuhan UniversityWuhanHubei 430072China Schoolof Bioengineering and Health Wuhan Textile UniversityWuhanHubei 430200China

A monolithic column-based mass spectrometry(MS)analysis kit was prepared for whole blood analysis with MS.The kit is disposable and can be used for purification,storage,transportation and direct analysis of whole blood.The kit mainly consists of a capillary for quantitative microsampling,a cation exchange monolithic column for purification and storage,and a syringe for loading sample.This kit is very friendly to various users that one can easily siphon the blood in the kit followed by rapid clean-up.We established a quantitative method using the kit with a limit detection as low as 0.33 nmol/L,and achieved more than five orders of magnitude enhancement in sensitivity compared to direct nanoelectrospray ionization MS analysis.The column can avoid analyte exposure to environment,which helps the storage of the sample for laboratory analysis.The relative standard deviation of immediate blood analysis and storage blood analysis within 10 d was less than 10%.This method has been successfully applied to the quantitative analysis of procainamide hydrochloride in 2μL rat blood.These results indicate that this disposable kit does have the potential to achieve highly sensitive quantitative MS analysis in biological samples,which is expected to become a cost-effective and powerful tool for in vitro diagnostics.

关键词： Clinical Ms Whole blood analysis Porous polymers Matrix isolation Trace analysis

A personality-guided affective brain-computer interface for implementation of emotional intelligence in machines

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Information Technology & Electronic Engineering 2022年第8期23卷 1158-1173页

作者： Shaojie LI wei LI Zejian XING Wenjie YUAN xiangyu wei Xiaowei ZHANG Bin HU School of Information Science and Engineering Lanzhou UniversityLanzhou 730099China

Affective brain-computer interfaces have become an increasingly important topic to achieve emotional intelligence in human–machine collaboration.However,due to the complexity of electroencephalogram(EEG)signals and the individual differences in emotional response,it is still a great challenge to design a reliable and effective model.Considering the influence of personality traits on emotional response,it would be helpful to integrate personality information and EEG signals for emotion recognition.This study proposes a personalityguided attention neural network that can use personality information to learn effective EEG representations for emotion recognition.Specifically,we first use a convolutional neural network to extract rich temporal and regional representations of EEG signals,and a special convolution kernel is designed to learn inter-and intra-regional correlations simultaneously.Second,inspired by the fact that electrodes within distinct brain scalp regions play different roles in emotion recognition,a personality-guided regional-attention mechanism is proposed to further explore the contributions of electrodes within a region and between regions.Finally,attention-based long short-term memory is designed to explore the temporal dynamics of EEG signals.Experiments on the AMIGOS dataset,which is a dataset for multimodal research for affect,personality traits,and mood on individuals and groups,show that the proposed method can significantly improve the performance of subject-independent emotion recognition and outperform state-of-the-art methods.

关键词： Electroencephalogram(EEG) Emotion recognition Attention mechanism Personality traits

维普期刊数据库

Formulation of Locational Marginal Electricity-carbon Price in Power Systems

在线全文

学校读者我要写书评

暂无评论

CSEE Journal of Power and Energy Systems 2023年第5期9卷 1968-1972页

作者： Junlong Li Yue Xiang Chenghong Gu wei Sun xiangyu wei Shuang Cheng Junyong Liu the Department of Electronic and Electrical Engineering University of BathBathUK the College of Electrical Engineering Sichuan UniversityChengduChina the School of Engineering University of EdinburghEdinburghUK

Decarbonisation of power systems is essential for realising carbon neutrality,in which the economic cost caused by carbon needs to be qualified.Based on the formulation of locational marginal price(LMP),this paper proposes a locational marginal electricity-carbon price(EC-LMP)model to reveal carbon-related costs caused by power consumers.A carbon-priceintegrated optimal power flow(C-OPF)is then developed to maximise economic efficiency of the power system considering the costs of electricity and carbon.Case studies are presented to demonstrate the new formulation and results demonstrate the efficacy of the EC-LMP-based C-OPF on decarbonisation and economy.

关键词： Carbon abatement cost carbon tariff locational marginal electricity-carbon price locational marginal price optimal power flow

Core-shell nanoparticle enhanced Raman spectroscopy in situ probing the composition and evolution of interfacial species on PtCo surfaces

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Nano Research 2024年第6期17卷 4687-4692页

作者： Jing Liu Han-Liang Zhong xiangyu Li Mu-Fei Yue wei-Min Yang Xueqiu You Jing-Hua Tian Yao-Hui Wang Jian-Feng Li Fujian Provincial Key Laboratory of Oceanic Information Perception and Intelligent Processing School of Ocean Information EngineeringJimei UniversityXiamen 361021China College of Chemistry Fuzhou UniversityFuzhou 350108China State Key Laboratory of Physical Chemistry of Solid Surfaces MOE Key Laboratory of Spectrochemical Analysis and InstrumentationiChEMCollege of Chemistry and Chemical EngineeringCollege of MaterialsCollege of EnergyXiamen UniversityXiamen 361005China Innovation Laboratory for Sciences and Technologies of Energy Materials of Fujian Province(IKKEM) Xiamen 361005China Shenzhen Research Institute of Xiamen University Shenzhen 518000China

The composition and evolution of interfacial species play a key role during electrocatalytic process.Unveiling the structural evolution and intermediate during catalytic process by in situ characterization can shed new light on the electrocatalytic reaction mechanism and develop highly efficient catalyst.However,directly probing the interfacial species is extremely difficult for most spectroscopic techniques due to complicated interfacial environment and ultra-low surface concentration.Herein,electrochemical core-shell nanoparticle enhanced Raman spectroscopy is utilized to probe the composition and evolution processes of interfacial species on Au@Pt,Au@Co,and Au@PtCo core-shell nanoparticle surfaces.The spectral evidences of interfacial intermediates including hydroxide radical(OH*),superoxide ion(O_(2)^(−)),as well as metal oxide species are directly captured by in situ Raman spectroscopy,which are further confirmed by the both isotopic experiment and density functional theory calculation.These results provide a mechanistic guideline for the rational design of highly efficient electrocatalysts.

关键词： in situ Raman core-shell nanoparticle PtCo interfacial species composition and evolution

A simplified noncryogenic strategy to transport mesenchymal stem cells: Potential applications in cell therapy and regenerative medicine

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Genes & Diseases 2024年第3期11卷 26-29页

作者： xiangyu Dong Yannian Gou Meichun Guo Jiamin Zhong Aohua Li Ailing Hao wei Zeng Rex C.Haydon Hue H.Luu Russell R.Reid Tongchuan He Yan Xu Jiaming Fan Ministry of Education Key Laboratory of Diagnostic Medicine And Department of Clinical BiochemistryCollege of Laboratory MedicineChongqing Medical UniversityChongqing 400016China Molecular Oncology Laboratory Department of Orthopaedic Surgery and Rehabilitation MedicineThe University of Chicago Medical CenterChicagoIL 60637USA Department of Interventional Neurology The First Dongguan Affiliated HospitalGuangdong Medical UniversityDongguanGuangdong 523475China Laboratory of Craniofacial Suture Biology and Development Department of Surgery Section of Plastic SurgeryThe University of Chicago Medical CenterChicagoIL 60637USA Institute of Sports Medicine Beijing Key Laboratory of Sports InjuriesPeking University Third HospitalBejing 100191China

With the rapid advances in stem cell research and po-tential cell-based therapies,there is an urgent need to develop safe and reliable cell transport strategies.Except for autologous stem cell-based therapies,allogeneic stem cell therapies and ex vivo genetically engineered cell therapies would require safe,efficient,and reliable cell preservation and transport methods.

关键词： transport methods simplified

Reduced clinical severity during 2022 Shanghai Spring epidemic of SARS-CoV-2 omicron BA.2 variant infection——an integrated account of virus pathogenicity and vaccination effectiveness

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

National Science Review 2024年第4期11卷 24-27页

作者： Xingyue Wu Yao Chen Kangli Cao Yao Shen Xueling Wu Yilin Yang Zhongshu Kuang Qingrun Li Zhenzhen Lu Yichen Jia Mian Shao Guorong Gu Xiangwei Wang Ye Yao Ying Wang Shaodie Chen Zhigao Yu wei wei Longfei Ding Lulu Lan Tianwen Gu xiangyu Long Jian Sun Lingyu Xing Jiayuan Shen Yi Han Yue Luo Sucheng Mu Mengna Lin Xiaoyan Zhang Rong Zeng Jianqing Xu Guoping Zhao Lihong Huang Zhenju Song Department of Emergency Medicine Clinical Center for Bio-Therapy Department of Biostatistics and Department of UrologyShanghai Key Laboratory of Lung Inflammation and Injury Zhongshan Hospital Fudan University Department of Respiratory and Critical Care Medicine Shanghai Pudong Hospital Fudan University Department of Pulmonology Renji Hospital Shanghai Jiao Tong University School of Medicine Key Laboratory of Systems Biology Center for Excellence in Molecular Cell Science Shanghai Institute of Biochemistry and Cell Biology Chinese Academy of Sciences Shanghai Public Health Clinical Center Fudan University Department of Biostatistics School of Public Health Key Laboratory of Public Health Safety of Ministry of Education and Shanghai Institute of Infectious Disease and Biosecurity School of Public Health Fudan University Shanghai Institute of Immunology Department of Immunology and Microbiology Shanghai Institute of Virology Key Laboratory of Cell Differentiation and Apoptosis of Chinese Ministry of EducationSchool of Medicine Shanghai Jiao Tong University State Key Laboratory of Genetic Engineering Fudan Microbiome CenterSchool of Life Sciences Fudan University Bio-Med Big Data Center Shanghai Institute of Nutrition and Health Chinese Academy of Sciences

The emergence of SARS-Co V-2 variant of concern omicron BA.2 led to high infection incidence rates as well as reduced severity of COVID-19 during the Shanghai Spring epidemic of 2022 [1]. We assessed the role of viral pathogenicity and the effectiveness of inactivated vaccines in a retrospective observational study including 10 258 confirmed COVID-19cases identified between April and June of 2022, from four COVID-19 designated hospitals located in two areas of Shanghai, namely the West Bund and the East Bund, where the populations’ demographic characteristics varied significantly (Table S1, Fig. S1).

关键词： REPLICATION

AIEgen configuration transition and aggregation enable dual prompt emission for single-component nondoped white OLEDs

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Aggregate

Aggregate 2024年第1期5卷 292-298页

作者： Jiasen Zhang Qiang wei wei Li Hao Chen xiangyu Zhu Yongqi Bai Nannan Fei Liang Cao Zujin Zhao Anjun Qin Ben Zhong Tang Ziyi Ge Zhejiang Provincial Engineering Research Center of Energy Optoelectronic Materials and Devices Ningbo Institute of Materials Technology and EngineeringChinese Academy of SciencesNingboP.R.China Center of Materials Science and Optoelectronics Engineering University of Chinese Academy of SciencesBeijingP.R.China Center for Aggregation-Induced Emission State Key Laboratory of Luminescent Materials and Devices South China University of Technology GuangzhouGuangzhouChina Shenzhen Institute of Aggregate Science and Technology School of Science and EngineeringThe Chinese University of Hong KongShenzhenChina

The dual emission(DE)feature in materials holds great potential to revolutionize the development of one-component system white organic lightemitting diodes(WOLEDs).However,the reported DE materials remain scarce owing to the formidable challenge of breaking Kasha’s rule and managing the intricate energy/charge transfer processes.Herein,we have introduced a groundbreaking DE AIEgen,2CzAn-TPE,which possesses a simple structure and undergoes Z-to-E isomerization and exhibits yellow and red fluorescence powders for pre-and post-sublimation,respectively.With relatively lower potential energy,Z-conformation((Z)-1,2-diphenyl-1,2-bis(4-(10-(9-phenyl-9H-carbazol-3-yl)anthracen-9-yl)phenyl)ethene)of 2CzAn-TPE can be readily transformed into E-conformation((E)-1,2-diphenyl-1,2-bis(4-(10-(9-phenyl-9Hcarbazol-3-yl)anthracen-9-yl)phenyl)ethene)via vacuum sublimation.The utilization of X-ray diffraction and grazing-incidence-wide-angle X-ray scattering techniques confirms the structural transformation,while the crystallographic analysis reveals the establishment of numerous intermolecular CH⋅⋅⋅πinteractions between the tetraphenylethene(TPE)moiety and both the anthracene and carbazole units.This allows a densely packed molecular arrangement,thereby offering propitious conditions for excimer generation in the E-conformation aggregated state.By utilizing the sublimated 2CzAn-TPE as an emitter,a nondoped one-component WOLED was prepared,exhibiting an exceptionally high external quantum efficiency(EQE)of 5.0%,which represents one of the highest performances among all one-componentWOLEDs.This research introduces a novel,simple,and efficient approach to realize highly efficient one-molecule WOLEDs.

关键词： aggregation-induced emission dual emission isomerization white OLED