Loading...
 首页  期刊介绍 期刊订阅 联系我们
 
最新录用  |  预出版  |  当期目录  |  过刊浏览  |  阅读排行  |  下载排行  |  引用排行  |  百年期刊

ISSN 1000-0585
CN 11-1848/P
Started in 1982
  About the Journal
    » About Journal
    » Editorial Board
    » Indexed in
    » Rewarded
  Authors
    » Online Submission
    » Guidelines for Authors
    » Templates
    » Copyright Agreement
  Reviewers
    » Guidelines for Reviewers
    » Online Peer Review
  Office
    » Editor-in-chief
    » Office Work
    » Production Centre
  • Table of Content
      , Volume 57 Issue 3 Previous Issue    Next Issue
    For Selected: View Abstracts Toggle Thumbnails
    COMPUTER SCIENCE AND TECHNOLOGY
    Case generation by constraints combinatorial testing
    BAI Xiaoying, HUANG Jun
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 225-233.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.001
    Abstract   PDF (1267KB) ( 1298 )
    The inputs parameters to software systems usually have various constraints correlated to each other. However, the input often violates the required constraints and constraint correlations. When there are many parameters with a large input domain, combinatorial testing can be used to reduce the test cost while ensuring test coverage. This paper defines a test adequacy criterion for constraint coverage using constrained combinatorial testing to detect conflicts caused by violating the constraints and constraint combinations. The system was applied to a member registration service for an online payment platform. Three typical combinatorial algorithms, OA (orthogonal array), IPO (in parameter order) and OFOT (one factor one time), were compared in forward their speeds, defect detection efficiency, and test case size using different configurations and parameter settings. The results showed that the OA algorithm was fast, produced fewer, more stable test cases with a reasonable defect detection efficiency, so this algorithm is good for iterative optimizations.
    Figures and Tables | References | Related Articles | Metrics
    Assessment of pure tone audiometry based on multiple judgments
    WU Yuhao, JIA Jia, ZHANG Xiulong, CAI Lianhong
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 234-239.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.002
    Abstract   PDF (1162KB) ( 924 )
    A auditory perception plays an important role in hearing assessments. A model was developed to transform amplitudes to sound pressure using acoustic parameters based on the auditory characteristics for automatic generation of standard pure tone test signals. The sounds were adjusted based on human auditory perception for a reliable algorithm to compute the hearing threshold based on multiple judgments. A correction algorithm for the hearing threshold is given for noise masking. A pure tone test was developed for an embedded platform to compare the hearing curves of normal people with standard hearing ranges. The results show that the system can quickly and easily check hearing changes, is reliable, and can be used for hearing screening.
    Figures and Tables | References | Related Articles | Metrics
    Speaker verification based on SVM and total variability
    GUO Wu, ZHANG Sheng, XU Jie, HU Guoping, MA Xiaokong
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 240-243.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.003
    Abstract   PDF (1117KB) ( 653 )
    The total variability factor extractor and the probability linear discriminant analysis (PLDA) algorithms have been the state-of-the-art for text-independent speaker verification. This study combines a support vector machine (SVM) with the PLDA. The low dimensional i-vectors of the total variability system are used as the inputs to the support vector machine, with the cosine kernel function used to achieve better discrimination. This method achieves considerable performance improvement with the PLDA system. Furthermore, the score fusion of the SVM with the PLDA give even better results. Tests were conducted on the female part of the interview section of the NIST 2012 core test corpus. The detection cost function (DCF) of the fusion system was reduced by 25.1% for common condition 1 and 25.2% for condition 3 compared with the best results for a single system.
    Figures and Tables | References | Related Articles | Metrics
    Task scheduling on a many-core processor for high-volume throughput applications
    XU Yuanchao, YANG Lu
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 244-249.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.004
    Abstract   PDF (1394KB) ( 931 )
    Big data applications with high-volume throughputs have become the most common applications in datacenters. The efficiencies of these applications running on traditional processors are very low for various reasons, one of which is the low-efficiency task scheduling. This paper presents a task scheduling framework that identifies program behavior and the running environment and then partitions the cores with hierarchical task scheduling though hardware and software co-design to reduce the negative effect of shared resource contention and improving the instruction cache hit rate using thread similarity. Tests show this algorithm improves performance by 20% on average over the legacy work-stealing scheduling algorithm.
    Figures and Tables | References | Related Articles | Metrics
    Speech-driven video-realistic talking head synthesis using BLSTM-RNN
    YANG Shan, FAN Bo, XIE Lei, WANG Lijuan, SONG Geping
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 250-256.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.005
    Abstract   PDF (1716KB) ( 1099 )
    This paper describes a deep bidirectional long short term memory (BLSTM) approach for speech-driven photo-realistic talking head animations. Long short-term memory (LSTM) is a recurrent neural network (RNN) architecture that is designed to model temporal sequences and their long-range dependencies more accurately than conventional RNNs. The deep BLSTM-RNN model is applied using a speaker's audio-visual bimodal data. The active appearance model (AAM) is used to model the facial movements with AAM parameters as the prediction targets of the neural network. This paper studies the impacts of different network architectures and acoustic features. Tests on the LIPS2008 audio-visual corpus show that networks with BLSTM layer(s) consistently outperform those having only feed-forward layers. The results show that the best network has a feed-forward layer inserted into two BLSTM layers with 256 nodes (BFB256) in the dataset. The combination of FBank, pitch and energy gives the best performance feature set for the speech-driven talking head animation task.
    Figures and Tables | References | Related Articles | Metrics
    Multilayer structure based lexicon optimization for language modeling
    Mijit Ablimit, Akbar Pattar, Askar Hamdulla
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 257-263.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.006
    Abstract   PDF (1191KB) ( 745 )
    An appropriate lexicon set must be selected as an important first step in developing large vocabulary continuous speech recognition (LVCSR) systems. The word unit is chosen as the lexicon basis to avoid word boundary detection problems. However, the lexicon basis selection is not as simple for the derivative morphological structure (e.g., agglutinative languages). Furthermore, there are no word boundaries in many languages such as Chinese and Japanese. This paper uses the Uyghur LVCSR system to analyze various particle based automatic speech recognition (ASR) systems with comparisons of the ASR results for various linguistic layers to develop a method to balance the advantages of two layer lexicons. The ASR results for the two layers are aligned and compared to analyze error patterns and extract samples as training data for the alternative selection method. Tests show that this method effectively improves the ASR accuracy with a small lexicon size.
    Figures and Tables | References | Related Articles | Metrics
    Social governance data integration and decision analysis method based on big data
    HONG Zhixu, CHEN Hao, CHENG Liang
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 264-269.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.007
    Abstract   PDF (1248KB) ( 1585 )
    This paper presents distributed data integration and visualization methods for distributed data integration for complex social governance data sets which have weak data analysis. The enhanced dynamic data description and Web visualization capabilities facilitate service-oriented intelligence society governance decision analyse using the large data processing mode to access, extract, integrate, and mine scattered data in different network routing databases. This paper presents a prototype system with trial evaluations. The results show that this method enhances real-time data integration, security, accuracy and data analysis visualization, and is more streamlined than the traditional model.
    Figures and Tables | References | Related Articles | Metrics
    Keyword extraction algorithms for emotion recognition from Uyghur text
    IMAM Seyyare, PARHAT Rayilam, HAMDULLA Askar, LI Zhijun
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 270-273.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.008
    Abstract   PDF (957KB) ( 622 )
    This paper describes sentiment classification research on Uyghur text using different keyword extraction methods to recognize common emotions like anger and happiness. The keywords expressing happiness and anger are extracted using the TextRank, sparse discriminant analysis (SDA) and sparse support vector machine (Sparse SVM) methods to train feature extraction and sentiment models. A sentiment text database was built by excerpting the anger and happiness sentiments from Uyghur movies and novels with several validation experiments based on those text databases. The tests show that the keyword extraction methods presented in this paper are effective for emotion classification from Uyghur sentences. The Sparse SVM method is robustness and has higher accuracy in recognition tests with a smaller number of keywords extracted.
    Figures and Tables | References | Related Articles | Metrics
    PHYSICS AND ENGINEERING PHYSICS
    Numerical simulations of leakage gas dispersion based on soil and atmosphere coupling
    WANG Yan, HUANG Hong, HUANG Lida, LI Yuntao
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 274-280.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.009
    Abstract   PDF (2501KB) ( 1210 )
    Most urban gas pipelines are buried underground. Small leaks can occur and are difficult to detect, which is a threat to urban public safety. Thus, gas leaks in buried gas lines need to be detected and assessed. The leaks can be located by analyzing the coupled gas dispersion in the soil and atmosphere. Different governing equations were used to model the dispersion in the soil and atmosphere in OpenFOAM. The flows were coupled through the mass flux leaving the ground surface. The models were used to investigate the impact of the leakage rate on the pressure and concentration distribution and the impact of the convective mass transfer coefficient on the concentration distribution and ground mass flux. The gas distribution in an urban street canyon was analyzed to show that the ground mass flux has little effect on the methane distribution in the street canyon. The model can be used as a reference for leak detection and assessments of buried gas pipelines.
    Figures and Tables | References | Related Articles | Metrics
    Experimental investigation of the thermal insulation and evaporative resistance of protective clothing on a thermal manikin in a hot environment
    FU Ming, WENG Wenguo, HAN Xuefeng
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 281-285,292.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.010
    Abstract   PDF (2900KB) ( 1465 )
    The thermal insulation and evaporative resistance of clothing are the two main parameters related to clothing thermal comfort and protective ability. These two parameters, that all usually only measured for normal or cold atmospheric conditions, are measured here using a thermal manikin, one a commonly used and accurate method. This paper presents measurements of the thermal insulation and evaporative resistance of two-layer and three-layer protective clothing using the thermal manikin "NEWTON" in normal and hot environments. The tests show that, the parallel resistance method gives lower overall thermal resistances than the serial resistance method in hot environments. The tests also show that the thermal insulation resistances at high temperatures are 30%~38% lower than at normal conditions. The results also show that evaporative resistance of the two-layer clothing in the hot environment is less than for the warm condition and that the evaporative resistance of the three-layer clothing based on the mass loss method at high temperatures is less than in the normal environment, while the evaporative resistance based on the heat loss method in the hot environment is larger than that in the normal environment.
    Figures and Tables | References | Related Articles | Metrics
    Speech enhancement algorithm that combines EEMD and K-SVD dictionary training
    GAN Zhenye, CHEN Hao, YANG Hongwu
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 286-292.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.011
    Abstract   PDF (1774KB) ( 979 )
    This paper presents a speech enhancement algorithm that combines the ensemble empirical mode decomposition (EEMD) algorithm and the K-singular value decomposition (K-SVD) dictionary-training algorithm. The EEMD algorithm is used to obtain the intrinsic mode function (IMF) components from noisy speech. The cross-correlations and autocorrelations of each IMF are calculated from the IMF components to filter out the noisy IMF components. The transition IMF components are again decomposed with EEMD to further remove the noisy component. The remained IMFs and transition IMFs are superimposed to generate the de-noised speech. An over-complete dictionary is then trained from the clean speech by the K-SVD dictionary training algorithm. The de-noised speech is then sparse decomposed with the over-complete dictionary to obtain the enhanced speech by recovering the speech signal from sparse coefficient vectors. Tests show that the algorithm achieves better de-noising than the traditional spectral subtraction, wavelet threshold de-noising and K-SVD dictionary-training algorithms for both low signal-to-noise ratio (SNR) and high SNR environments.
    Figures and Tables | References | Related Articles | Metrics
    ELECTRICAL ENGINEERING
    Priority weighted rate control algorithm in aeronautical Ad hoc networks
    GAO Xiaolin, YAN Jian, LU Jianhua
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 293-298.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.012
    Abstract   PDF (1105KB) ( 721 )
    Current media access control (MAC) protocols for aeronautical Ad hoc networks can not achieve fair channel resource utilization. This paper describes a priority weighted rate control (PWRC) algorithm, which uses the fair queuing model in the MAC layer. The traffic speed reduction factor is determined by its priority and weight. The PWRC guarantees the quality of service (QoS) while the channel resource is fairly utilized. Simulations indicate that the traffic fairness index with the same priority is closer to the upper band 1 than with existing MAC protocols. The channel resource utilization is improved about 40% and the end-to-end delay and jitter are reduced.
    Figures and Tables | References | Related Articles | Metrics
    Perception of Cantonese tones
    XIE Chenwei, SHI Feng, WEN Baoying
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 299-305.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.013
    Abstract   PDF (1923KB) ( 736 )
    The perceptual of Cantonese speeds tones was investigated for the perceptual of three tones on identification and discrimination experiments. This research used a Cantonese natural language monosyllabic and disyllabic corpus as the speech material with identification and discrimination experiments based on the "Diagonal/Half Space Test Method". The monosyllable tests showed that the three groups of high and low tones have their own perceptual space with the perceptual space of the low tones substantially overlap pins that of the mid level tones, which is consistent with the acoustic space of these tones. The bisyllable tests indicate that the perceptual category is certain with a clear categorical boundary. The different tones of the reference words and the first part or the last part of disyllabic words affect the categorical boundary location.
    Figures and Tables | References | Related Articles | Metrics
    CIVIL ENGINEERING
    Influence of ferro-nickel slag powder on the sulfate attack resistance of concrete
    WANG Qiang, SHI Mengxiao, ZHOU Yuqi, YU Chenghang
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 306-311.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.014
    Abstract   PDF (1247KB) ( 1038 )
    The influence of ferro-nickel slag powder on the sulfate attack resistance of concrete and the corresponding mechanism were investigated by measuring the porosity and charge passed through the concrete and the compressive strength loss below and after the concrete experienced a dry-wet cycle in a sodium sulfate solution and immersion in a high temperature sodium sulfate solution. The results show that the effect of the two kinds of sulfate attacks were the same. The connected porosity of the concrete increases and its sulfate attack resistance decreases with increasing electric furnace ferro-nickel slag powder content. The connected porosity of the concrete decreases, the chloride ion penetrability decreases and its sulfate attack resistance increases with increasing blast furnace ferro-nickel slag powder with mass fractions less 30%. The sulfate attack resistance of concrete is enhanced less by adding blast furnace ferro-nickel slag powder than by adding blast furnace slag power especially at high powder mass fractions.
    Figures and Tables | References | Related Articles | Metrics
    AUTOMATION
    Personnel location oriented global mine tunnel network modeling
    LI Chunwen, XING Zhipeng, LU Sicong
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 312-317.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.015
    Abstract   PDF (1178KB) ( 777 )
    Tunnels are the core of mine projects and tunnel network modeling is an important part of digital mine construction. The mine underground personnel position system is one of the "six major systems" for reducing risks in mines.Coal mines have multi-level, intersecting tunnel networks that complicate underground personnel positioning systems. This paper used some assumptions to develop a method for single level mine roadway network global modeling. The method uses an arc-node continuous model of the roadway network and the arc-node discrete sampling model with the fitting and transformations between models. This modeling method provids a framework for the design of mine personnel location systems and other large-scale dynamic monitoring systems.
    Figures and Tables | References | Related Articles | Metrics
    HYDRAULIC ENGINEERING
    Digital filter baseflow separation method based on a master recession curve
    ZHOU Xing, SHEN Chen, NI Guangheng, HU Hongchang
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 318-323,330.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.016
    Abstract   PDF (2339KB) ( 1293 )
    Baseflow separation is essential in hydrology, water resources and water ecology. This study compares several baseflow separation methods with a digital filter baseflow separation method based on a master recession curve then developed to reduce the parameter uncertainty. A study of the Nu river in China shows that this digital filter method gives smoother baseflows than the BFI method and the HYSEP method; however, it is sensitive to model the parameters. Reasonable baseflows were obtained by using a recession curve to determine the digital filter parameters. The results show that the baseflow in the Nu River accounts for 39 to 55 percent of the total runoff in different sub-basins and that the inter-annual variation of the baseflow is relatively small.
    Figures and Tables | References | Related Articles | Metrics
    MECHANICAL ENGINEERING
    Multivariable coating thickness distribution model for robotic spray painting
    WANG Guolei, YI Qiang, MIAO Dongjing, CHEN Ken, WANG Liqiang
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 324-330.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.017
    Abstract   PDF (1408KB) ( 1092 )
    A multivariable robotic spray painting model was developed for a range of painting parameters to improve the restricted traditional model. A β distribution based coating growth rate function was used with a coating thickness distribution formula then deduced from the integral of the growth rate function. The maximum coating growth rate was related to the paint flow rate and painting distance with the paint flow rate related to the painting air pressure and painting width from experimental data. Then, a generalized coating thickness distribution model was developed with five painting parameters as independent variables by substituting these relations into the coating thickness distribution equations. The model was validated through experiments with the results showing that it can predict the coating thickness distribution for various painting parameters with an average forecasting error of less than 4.3%.
    Figures and Tables | References | Related Articles | Metrics
    ENVIRONMENTAL SCIENCE AND ENGINEERING
    Water conservation approaches in printing and dyeing production processes based on wastewater treatment and graded reuse
    ZHOU Lu, XU Kun, QIU Zhaojing
    Journal of Tsinghua University(Science and Technology). 2017, 57 (3): 331-336.   DOI: 10.16511/j.cnki.qhdxxb.2017.26.018
    Abstract   PDF (1431KB) ( 558 )
    A water conservation database was developed to improve the wastewater reuse rate in printing and dyeing production for batch production processes having drastic fluctuations in wastewater flowrate and pollutant concentrations. The water conservation data was analyzed using a real-time parallel computations based on the textile company's information collection system. The wastewater quality parameters such as the chemical oxygen demand (COD) and conductivity were monitored during wastewater treatment to evaluate the reuse potential for water resource management. A scheduling and reuse system using the production processes as the analysis object was developed to improve the total wastewater discharge and reuse rates. A typical knit cotton dyeing enterprise was used for a case analysis. The results show that the system reduces water consumption by 38.72% and also reduces the quality fluctuations and wastewater flowrate. This has a positive effect on the operating stability of the water treatment processes.
    Figures and Tables | References | Related Articles | Metrics
  News More  
» aaa
  2024-12-26
» 2023年度优秀论文、优秀审稿人、优秀组稿人评选结果
  2023-12-12
» 2022年度优秀论文、优秀审稿人、优秀组稿人评选结果
  2022-12-20
» 2020年度优秀论文、优秀审稿人评选结果
  2021-12-01
» aa
  2020-11-03
» 2020年度优秀论文、优秀审稿人评选结果
  2020-10-28
» 第十六届“清华大学—横山亮次优秀论文奖”暨2019年度“清华之友—日立化成学术交流奖”颁奖仪式
  2020-01-17
» a
  2019-01-09
» a
  2018-12-28
» a
  2018-01-19


  Links More  



Copyright © Journal of Tsinghua University(Science and Technology), All Rights Reserved.
Powered by Beijing Magtech Co. Ltd