检索结果-南通市图书馆

同方期刊数据库

Aligning enhanced feature representation for generalized zero-shot learning

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2025年

作者： Zhiyu FANG Xiaobin ZHU Chun YANG Hongyang ZHOU Jingyan QIN Xucheng YIN School of Computer&communication Engineering University of Science and Technology Beijing

Abstract Constructing an effective common latent embedding by aligning the latent spaces of cross-modal variational autoencoders （VAEs） is a popular strategy for generalized zero-shot learning （GZSL）.However,due to the lack of fine-grained instance-wise annotations,existing VAE methods can easily suffer from the posterior collapse *** this paper,we propose an innovative asymmetric VAE network by aligning enhanced feature representation （AEFR） for *** from general VAE structures,we designed two asymmetric encoders for visual and semantic observations and one decoder for visual ***,we propose a simple yet effective gated attention mechanism （GAM） in the visual encoder for enhancing the information interaction between observations and latent variables,alleviating the possible posterior collapse problem *** addition,we propose a novel distributional decoupling-based contrastive learning （D2-CL） to guide learning classification-relevant information while aligning the representations at the taxonomy level in the latent representation *** experiments on publicly available datasets demonstrate the state-of-the-art performance of our *** source code is available at https://***/seeyourmind/AEFR.

关键词：

Fast degree-distribution optimization for BATS codes

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2017年第10期60卷 124-138页

作者： Huakai ZHAO Shenghao YANG Guinian FENG beijing Satellite Navigation Center The Chinese University of Hong Kong beijing Institute of tracking and communication technology

Batched sparse(BATS) codes have been proposed for communication through networks with packet loss. BATS codes include a matrix generalization of fountain codes as the outer code and random linear network coding at the intermediate network nodes as the inner code. BATS codes, however, do not possess a universal degree distribution that achieves an optimal rate for any distribution of the transfer matrix ranks. Therefore,it is important to have a fast degree-distribution optimization approach for finite-length BATS codes. In this paper, we propose the concept of batch release probability(BRP), and demonstrate some characteristics of BRPs from the degree distributions achieving nearly optimal performance. Based on these BRP characteristics,we propose a novel degree-distribution optimization approach that achieves the similar decoding performance with a much shorter optimization time, compared with the previous approach. Moreover, the universality of BRPs observed in this paper can further simplify the degree-distribution optimization of BATS codes.

关键词： network coding batched sparse(BATS) code finite-length analysis degree distribution batch release probability optimization

Numerical Simulation of Radiation Transport to Improve Microalgae Cultivation in an Air-Lift Photobioreactor

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 Thermal Science 2025年第1期34卷 62-76页

作者： ZHANG Xianbin LI Mingjia HUNG Tzuchen ZHANG Zixun Key Laboratory of Thermo-Fluid Science and Engineering of Ministry of Education School of Energy and Power EngineeringXi'an Jiaotong University School of Mechanical Engineering Beijing Institute of Technology Department of Mechanical Engineering Taipei University of Technology

A comprehensive numerical model is developed to simulate the growth of microalgae under light/dark cycling *** purpose of this study is to predict the growth rate of Chlorella vulgaris cultivated in photobioreactors(PBRs) in order to improve the light conditions for microalgae and enhance the photosynthetic *** fluid dynamics(CFD) is used to simulate its internal hydrodynamic *** Lagrangian method is employed to track the movement of microalgae *** radiative transfer equation(RTE)is used to obtain light intensity *** combination of light radiation field and microalgae cell motions is used to construct the light history and they are integrated into the model of the photosynthetic units(PSU) to calculate the microalgae growth *** numerical results demonstrate that enhanced light/dark cycling frequency with ordered mixing can promote efficient microalgae *** effect of the vortex flow field generated by the baffles in an air-lift PBR is analyzed for increasing microalgae growth *** using the 1:1 baffle spacing,the biomass production of microalgae is increased by 41.8% compared to the original PBR.

关键词： light distribution particle tracking photobioreactor structure optimization

Deep learning-based software engineering: progress,challenges, and opportunities

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2025年第1期68卷 57-144页

作者： Xiangping CHEN Xing HU Yuan HUANG He JIANG Weixing JI Yanjie JIANG Yanyan JIANG Bo LIU Hui LIU Xiaochen LI Xiaoli LIAN Guozhu MENG Xin PENG Hailong SUN Lin SHI Bo WANG Chong WANG Jiayi WANG Tiantian WANG Jifeng XUAN Xin XIA Yibiao YANG Yixin YANG Li ZHANG Yuming ZHOU Lu ZHANG School of Journalism and communication Sun Yat-sen University School of Software technology Zhejiang University School of Software Engineering Sun Yat-sen University School of Software Dalian University of Technology School of Computer Science and technology Beijing Institute of Technology Key Laboratory of High Confidence Software Technologies (Peking University) Ministry of EducationSchool of Computer Science Peking University State Key Laboratory for Novel Software technology Nanjing University School of Computer Science and Engineering Beihang University Institute of Information Engineering Chinese Academy of Sciences School of Computer Science Fudan University State Key Laboratory of Complex & Critical Software Environment (CCSE) School of Software Beihang University School of Computer and Information technology Beijing Jiaotong University School of Computer Science and technology Harbin Institute of Technology School of Computer Science Wuhan University Huawei Technologies

Researchers have recently achieved significant advances in deep learning techniques, which in turn has substantially advanced other research disciplines, such as natural language processing, image processing, speech recognition, and software engineering. Various deep learning techniques have been successfully employed to facilitate software engineering tasks, including code generation, software refactoring, and fault localization. Many studies have also been presented in top conferences and journals, demonstrating the applications of deep learning techniques in resolving various software engineering tasks. However,although several surveys have provided overall pictures of the application of deep learning techniques in software engineering,they focus more on learning techniques, that is, what kind of deep learning techniques are employed and how deep models are trained or fine-tuned for software engineering tasks. We still lack surveys explaining the advances of subareas in software engineering driven by deep learning techniques, as well as challenges and opportunities in each subarea. To this end, in this study, we present the first task-oriented survey on deep learning-based software engineering. It covers twelve major software engineering subareas significantly impacted by deep learning techniques. Such subareas spread out through the whole lifecycle of software development and maintenance, including requirements engineering, software development, testing, maintenance, and developer collaboration. As we believe that deep learning may provide an opportunity to revolutionize the whole discipline of software engineering, providing one survey covering as many subareas as possible in software engineering can help future research push forward the frontier of deep learning-based software engineering more systematically. For each of the selected subareas,we highlight the major advances achieved by applying deep learning techniques with pointers to the available datasets i

关键词： deep learning software engineering software benchmark software artifact representation survey

Relative position determination between Chang'E-3lander and rover using in-beam phase referencing

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2015年第9期58卷 161-170页

作者： ZHOU Huan LI Hai Tao DONG GuangLiang beijing Institute of tracking and Telecommunications technology

Chang'E-3 is China's first lunar surface exploration mission, as a part of the second phase of China's Lunar Exploration Program(CLEP). It landed successfully in the northwest of Mare Imbrium on December 14,2013, incorporating two probes, a lander and a rover named Yutu. The execution of the cooperative science tasks required an accurate determination of the relative position between the lander and the rover. In-beam phase referencing experiments have been conducted with the Chinese VLBI Network(CVN) to accurately image the rover in reference to the stationary lander in order to obtain their offsets in Right Ascension and Declination,which were then transformed into the lunar local coordinate system to acquire their relative position on the lunar surface. The observations and the data processing procedures are described in this paper. Images of the rover in six epochs between December 15 and 21, 2013 were made, resulting in accurate positions of the rover that are consistent with the visual localization approach. Through measurement of the locations of the rover's omnidirectional antenna with respect to the lander's directional antenna before their separation, and the recognition of the rover's turning actions during its exploratory moves, we demonstrate that the accuracy of relative position determined in this way could reach the meter level and the resolution is better than 1 m.

关键词： Chang’E-3 relative position determination in-beam phase referencing Chinese VLBI Network lunar exploration

Initial result of the Chinese Deep Space Stations' coordinates from Chinese domestic VLBI experiments

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2017年第1期60卷 170-176页

作者： Guangliang DONG Dezhen XU Haitao LI Huan ZHOU beijing Institute of tracking and Telecommunications technology

China's Lunar Exploration Program(CLEP) prompted the design and construction of the globally distributed Chinese Deep Space Network(CDSN). This network consists of Jiamusi and Kashi stations in China, and Zapala station in Argentina. However, the positions of Jiamusi and Kashi are not accurate enough for future Chinese deep space missions, and geodetic Very Long Baseline Interferometry(VLBI) is the most effective way to determine their positions. Since the CDSN stations are equipped with narrow-band receivers,they cannot participate in current international VLBI sessions in which wide-band frequencies are utilized. Thus a cooperative geodetic program of the CDSN and Chinese VLBI Network(CVN, the VLBI tracking subsystem of the CLEP) was initiated to determine their positions, in which specially designed frequencies can be utilized,and some CVN stations can act as position reference stations owing to their precise positions from long-term international VLBI observations. Primary results have been obtained from the CDSN–CVN combined domestic VLBI experiments from September 28, 2014, through December 10, 2015. The positions of Jiamusi and Kashi are determined to be better than 10-mm precision in the X, Y, and Z directions, which are improved by a factor of approximately 20 over their a priori values.

关键词： VLBI Chinese Deep Space Network Chinese VLBI Network station position determination terrestrial reference frame

Improved position estimates for the Chinese Deep Space Station Kashi derived by geodetic very long baseline interferometry

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2016年第9期59卷 1346-1351页

作者： XU DeZhen DONG GuangLiang ZHOU Huan LI HaiTao beijing Institute of tracking and Telecommunications technology Beijing 100094 China

A dedicated 24 h S/X dual-band geodetic very long baseline interferometry(VLBI) experiment was conducted in January 2015 with the goal of improving the position estimates for the Chinese Deep Space Station Kashi. Previously, the position estimates had been only accurate to ~20 cm, which is insufficient for future Chinese deep space explorations. The experiment design and data reduction are described with special emphasis on the limited frequency ranges of Kashi for bandwidth synthesis. A narrowed multi-band delay search window based on post-fit residuals was utilized to resolve the sub-ambiguities due to the drop of a frequency channel in fringe fit, which saved ~22% of the observations from the affected baseline. Final position estimates of Kashi were obtained from the global solution by using more than 5300 international VLBI sessions from August 1979 to September 2015, and estimates were found to be accurate to about 10, 25, and 20 mm in the X, Y, and Z components. Various statistical tests were run, and the estimates and precisions are believed to be reliable.

关键词： VLBI Chinese Deep Space Station station position determination bandwidth synthesis quantitative efficiency

维普期刊数据库

New high-accuracy spacecraft VLBI tracking using high data-rate signal:A demonstration with Chang'E-3

在线全文

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2016年第4期59卷 558-564页

作者： ZHOU Huan XU DeZhen CHEN ShaoWu LI HaiTao DONG GuangLiang beijing Institute of tracking and Telecommunications technology Beijing 100094China

As the scientific data volume in deep-space exploration rapidly grows,spacecraft heavily relies on high data-rate signals that span several megahertz to transmit data back to *** high data-rate signals for high-accuracy radiometric interferometry can simultaneously deal with data transmission and spacecraft *** demonstrate very long baseline interferometry(VLBI)tracking of the Chang’E-3 lander and rover to determine their relative lunar-surface position using downlink high data-rate signals.A new method based on the VLBI phase-referencing technique is proposed to obtain the differential phase delay,which is much more accurate than the differential group delay acquired by conventional VLBI *** systemic errors among different signal channels have been well calibrated using the new *** data from the Chang’E-3mission were then processed,and meter-level accuracy positions of the rover with respect to the lander have been *** demonstration shows the feasibility of high-accuracy radiometric interferometry using high data-rate *** method proposed in this paper can also be applied to future deep-space navigation.

关键词： spacecraft VLBI tracking high data-rate signal differential phase delay Chang'E-3 deep-space navigation

维普期刊数据库

A Novel Kind of Transverse Micro-Stack High-Power Diode Bars

在线全文

学校读者我要写书评

暂无评论

Chinese Physics Letters 2008年第4期25卷 1284-1286页

作者：张蕾崔碧峰李建军郭伟玲王智群沈光地 Opto-Electronic technology Laboratory Beijing University of Technology Beijing 100022 beijing Institute of tracking and Telecommunication technology Beijing 100094

Novel transverse micro-stack semiconductor laser bars are put forward to improve the output optical power of semiconductor laser bars at low injection current. More importantly, the novel laser bars have a coupled large optical cavity, which can overcome the problem of catastrophic optical damage and improve light beam quality due to the coherently coupled emitting along the transverse direction. The micro-stack tunnel regeneration tri-active region laser structure was grown by metal organic chemical vapour deposition. For a weakly coupled uncoated device, the optical power exceeds 60 W under 50 A driving current and the slope efficiency reaches 1.55 W/A. Further experiments show that the perpendicular divergence of 23° is achieved from transverse strongly coupled devices.

关键词： SEMICONDUCTOR-LASERS