A SPEAKER ADAPTABLE VERY LOW BIT RATE SPEECHCODER BASED ON HMM

IntroductionPhonetic and segmentvocoders are popular tech-niques used in speech coding at rates on the or-der of 1 0 0 bit/s.These coders are based onrecognition/synthesis technique.They use somekind of speech recognition technique to decom-pose speech signals to sequence of speech units,such as phonetic units or speech segment units,and then transmit indexes and durations of therecognized units.The decoders receive this in-formation,and synthesize speech by concatenat-ing typical instances of…     

Fuzzy C-Means Clustering Based Phonetic Tied-Mixture HMM in Speech Recognition

A fuzzy clustering analysis based phonetic tied-mixture HMM (FPTM) was presented to decrease parameter size and improve robustness of parameter training. FPTM was synthesized from state-tied HMMs by a modified fuzzy C-means clustering algorithm. Each Gaussian codebook of FPTM was built from Gaussian components within the same root node in phonetic decision tree. The experimental results on large vocabulary Mandarin speech recognition show that compared with conventional phonetic tied-mixture HMM and state-tied HMM with approximately the same number of Gaussian mixtures, FPTM achieves word error rate reductions by 4.84% and 13.02% respectively. Combining the two schemes of mixing weights pruning and Gaussian centers fuzzy merging, a significantly parameter size reduction was achieved with little impact on recognition accuracy.     

结合人体运动特征的行为识别

人体运动具有马尔可夫性质,即当前状态只受前一状态的影响.目前为止,用于人体行为识别的隐马尔可夫模型（HMM）大多使用的是全连接结构（Full-Connected structure）,并且没有把状态数目的选取和状态转移条件与人体运动特性间的关系作为研究重点.本文针对这种关系提出了基于“从左到右三状态半连接HMM”的人体行为识别方法,为每个状态的输出概率引入了权重的概念.实验表明,该方法能够在降低运算复杂度的同时,提高行为识别率,从而证明了人体运动特性分析在HAR领域中的应用价值.     

Object tracking and tracing: Hidden semi-Markov model based probabilistic location determination

The enhancement of radio frequency identification (RFID) technology to track and trace objects has attracted a lot of attention from the healthcare and the supply chain industry. However, RFID systems do not always function reliably under complex and variable deployment environment. In many cases, RFID systems provide only probabilistic observations of object states. Thus, an approach to predict, record and track real world object states based upon probabilistic RFID observations is required. Hidden Markov model (HMM) has been used in the field of probabilistic location determination. But the inherent duration probability density of a state in HMM is exponential, which may be inappropriate for modeling of object location transitions. Hence, in this paper, we put forward a hidden semi-Markov model (HSMM) based approach for probabilistic location determination. We evaluated its performance comparing with that of the HMM-based approach. The results show that the HSMM-based approach provides a more accurate determination of real world object states based on observation data.     

驾驶人“感知-决策-操控”行为模型

为准确模拟驾驶人跟车行为，提出基于隐马尔可夫模型(Hidden Markov Model，HMM)的驾驶人“感知-决策-操控”行为模型。建立描述驾驶意愿的HMM模型，模拟驾驶人感知过程，获得期望的车间距；预测模块模拟驾驶人根据交通环境和自身生理、心理状态预测车辆未来轨迹，即决策过程；优化模块描述驾驶人为使预测的车辆轨迹跟踪上期望的车辆间距而采取的操控汽车的执行动作，即操控过程。上述3个模块的滚动过程实现了对驾驶人跟车行为的模拟。利用自然驾驶数据进行算例分析，结果表明，本文模型预测车间距平均误差仅为1.47%，证明了所建模型的有效性及准确性。本文为驾驶行为建模方法的理论研究和应用拓宽了思路。     

Human-imitation recognition algorithm based on multi-character

A multi-character recognition method based on hidden Markov model (HMM) was presented. The method can reduce the calculation load of correlation and improve recognition accuracy compared with single-character recognition in video. The characteristics used for recognizing include the shape character, the color character, the texture character and so on. Even our human being generally uses these characteristics to recognize objects in practice. A recognition experiment of 17 fishes was carried out in the paper. The experimental results demonstrate the high veracity of the multi-character recognition algorithm. Together with the tracking process, it can handle dynamic objects, so the multi-character recognition is more like the human recognition, and has great application value.     

隐Markov模型在智能交通领域中应用

隐马尔科夫模型（HMM）是一种广泛使用的统计模型,在生命科学、信息科学、智能交通等领域得到广泛的应用。简述隐Markov模型的基本原理、核心问题及其改进模型,详细阐述隐马尔科夫模型在交通流阻塞智能检测、车辆碰撞预警和驾驶员意图识别等智能交通领域应用情况,并对其应用前景进行展望。     

自适应期望跟车间距和行为习惯的驾驶人跟驰模型

为满足智能车辆的个性化需求，提高智能车辆人-机交互协同的满意度和接受度，构筑双层驾驶人跟驰模型框架，提出自适应驾驶人期望跟车间距和行为习惯的个性化驾驶人跟驰模型。首先，提取个体驾驶人跟驰均衡状态的数据，采用高斯混合和概率密度函数(Gaussian Mixture Model and Probability Density Function, GMM-PDF)建立第 1 层模型，即驾驶人期望跟车距离模型。然后，将期望跟车距离参数引入模型，基于高斯混合-隐马尔可夫方法(Gaussian Mixture Model and Hidden Markov Model, GMM-HMM)学习驾驶习性，建立第2层模型预测加速度，即个性化驾驶人跟驰模型。其次，研究不同高斯分量个数对模型效果的影响，对比双层模型与 Gipps 模型、最优间距模型(Optimal Distance Model, ODM)、单层模型及通用模型的性能。最后，8位被试驾驶人的自然驾驶行为数据验证结果表明：高斯分量数量与模型性能存在一定的正相关性；在最优高斯分量数量下，8位被试驾驶人在训练集上预测误差均值为0.101 m·s-2，在测试 集上为0.123 m·s-2；随机选取其中1位驾驶人的2个跟车片段数据进行模型计算，结果显示，加速度的平均误差绝对值分别为0.087 m·s-2和0.096 m·s-2，预测效果优于Gipps模型、ODM模型、单层 模型及通用模型30%以上，与驾驶人实际跟驰行为的吻合度更高。     

EFFECTS OF DESENSITIZATION AND REBOUND TO ADENOSINE ON ACTION POTENTIAL AND CONTRACTILITY IN ATRIAL CELLS IN GUINEA-PIGS

Ａｄｅｎｏｓｉｎｅ (Ａｄｏ)ｉｓａｎｅｎｄｏｇｅｎｏｕｓｐｕｒｉｎｅｎｕ ｃｌｅｏｓｉｄｅｔｈａｔｒｅｇｕｌａｔｅｓｍａｎｙｐｈｙｓｉｏｌｏｇｉｃａｌａｎｄｐａｔｈｏ ｐｈｙｓｉｏｌｏｇｉｃａｌｐｒｏｃｅｓｓｅｓｉｎｂｏｄｙ .Ｔｈｅｅｆｆｅｃｔｓｏｎｃａｒｄ ｉｏｖａｓｃｕｌａｒｓｙｓｔｅｍａｒｅｅｓｐｅｃｉａｌｌｙｏｂｖｉｏｕｓ .Ｐｅｏｐｌｅｆｏｕｎｄｔｈａｔａｃｅｔｙｌｃｈｏｌｉｎｅ (ＡＣｈ)ａｎｄａｄｅｎｏｓｉｎｅｈａｄｔｈｅｎｅｇａｔｉｖｅｉｎｏｔｒｏｐｉｃ ,ｎｅｇａｔｉｖｅｃｈｒｏｎｏｔｒ…     

二色补血草止血作用的药理研究

二色补血草水煎剂15g(生药)/kg,乙醇提取液15g(生药)/kg灌胃,均使大白鼠的出血时间明显缩短,对凝血时间无明显影响。二色补血草醇提取液20g(生药)/kg灌胃,对大白鼠实验性动脉血栓形成有明显促进作用。因为动脉血栓形成主要与血小板的粘附和聚集有关。结果提示二色补血草止血作用机理可能与其促进血小板粘附和聚集作用有关。     

胶束荧光法对兔血浆中胺碘酮药代动力学的研究

利用具有高灵敏度的胶束荧光法,对胺碘酮的血药浓度进行了监测;使用SLS为胶束试剂,以国产930型荧光光度计为主要分析仪器,得到系列药代动力学参数:t_(1/2)(α)为1.18h,t_(1/2)(β)为40.75h,K_(21)为0.278h~(-1),K_(10)为0.036h~(-1),K_(12)为0.291h~(-1)。该法的线性范围为2×10~(-9)～8×10~(-6)mol/L,检测限为1.3×10~(-9)mol/L,平均回收率为99.93%。     

纤溶酶原激活物抑制物1与神经母细胞瘤

目的　研究纤溶酶原激活物抑制物 1 (PAI 1 )在神经母细胞瘤中的表达情况 ,评估其在肿瘤浸润转移中的作用。方法　应用免疫组化ABC法检测 42例神经母细胞瘤中PAI 1的表达 ;用RT PCR方法检测患儿骨髓和外周血中的神经蛋白基因产物 9 5 (PGP9 5 )mRNA ,确定瘤细胞的转移程度。结果　PAI 1阳性染色定位于肿瘤细胞和少许血管内皮细胞膜以及部分肿瘤细胞胞浆。PAI 1在神经母细胞瘤中的阳性表达率为 71 4% (3 0 /42 ) ,在进展期肿瘤阳性表达率是 85 7% (2 4 /2 8) ,局灶期肿瘤是 42 9% (6 /1 4) ,两者相比差异有显著性 (P <0 0 1 ) ;在预后不良组织型 (UH)病例中 ,PAI 1阳性率是 91 7% (2 2 /2 4 ) ,与预后良好型 (FH)病例的 44 4%(8/1 8)相比 ,有显著性差异 (P <0 0 1 )。在PAI 1阳性病例组 ,PGP9 5阳性检出率为 6 0 % (1 8/3 0 ) ,PAI 1阴性病例组为 8 3 % (1 /1 2 ) ,两者相比差异亦有显著性 (P <0 0 1 )。结论　PAI 1在神经母细胞瘤的浸润转移过程中发挥重要的作用。     

一类二型Fuchsian群的局部化性质

为了讨论整数的整二次型表示,引进整二次型的θ-级数,并发现θ-级数与模群的自守形式有紧密的联系.Fuchsian群及其自守形式是模群与模形式的重要延伸.对于一类二型Fuchsian群H(4q),这是G(然)的一个子群.众所周知,G(√q)中的元素要落H(√q)中需要有诸多限制.简化这些限制条件是很有意义的.利用p -局...     

Capillary electrophoresis with inhibited electrochemiluminescent detection for the trace analysis of epinephrine and dopamine

In this paper, a novel electrochemiluminescent (ECL) detection approach was developed for highly sensitive detection of ECL inhibitors based on the ECL inhibition of Ru(bpy)32+/2-(Dibutylamino)ethanol (DBAE) system. A microfluidic ECL detection cell was fabricated to couple with the capillary electrophoresis system, the electrochemical system and the postcolumn injection system. Both Ru(bpy)32+ and DBAE solutions were injected directly to the working electrode surface by a micro-infusion system to obtain a high and stable ECL signal. The performance of this setup was demonstrated by the analysis of two typical ECL inhibitors, dopamine and epinephrine. Under the optimal conditions, the limit of detection (LOD) for dopamine and epinephrine was 50nM and 5nM respectively. The proposed method was also successfully used for the trace analysis of dopamine and epinephrine in human serum samples.     

罗红霉素分散片的人体药物动力学及相对生物利用度

目的　研究罗红霉素分散片在健康人体内的药物动力学和相对生物利用度。方法　 1 2名健康男性志愿者 ,采用随机交叉试验设计 ,单剂量口服国产罗红霉素分散片和胶囊 30 0mg ,用微生物检定法测定不同时间的血药浓度。经 3P97程序拟合 ,计算药物动力学参数。结果　罗红霉素分散片和胶囊血药浓度 -时间曲线均符合二室模型 ,主要药物动力学参数 :T1 / 2 β分别为(1 4.76± 2 .77)h和 (1 4.92± 3.1 2 )h ,Tmax分别为 (1 .38± 0 .2 3)h和 (1 .63± 0 .2 3)h ,Cmax分别为(8.37± 0 .66)mg/L和 (8.2 3± 0 .63)mg/L ,AUC0→∞ 分别为 (91 .84± 1 1 .90 )h·mg/L和 (88.55± 1 0 .70 )h·mg/L。经统计学分析 ,两制剂的Cmax和AUC0→∞ 均无显著性差异 (P >0 .0 5) ,罗红霉素分散片的相对生物利用度为 (1 0 3.75± 5.58) %。结论　两制剂具有生物等效性。     

雷帕霉素对小鼠H_(22)肝癌细胞生长增殖的影响

目的 研究雷帕霉素(rapamycin, RPM)对H_(22)肝癌细胞生长增殖的影响.方法 体外培养小鼠H_(22)肝细胞癌株,分别与RPM、CsA、FK506和5-Fu共同孵育48 h,进行MTT实验.流式细胞仪测定各组H_(22)细胞的细胞周期变化,ELISA法测定各组上清液中VEGF含量.体内实验建立H_(22)肝细胞癌皮下移植瘤模型,同时完成C57BL/6→BALB/c小鼠异体皮肤移植模型.给予RPM、CsA、FK506和5-Fu灌胃,观察皮片存活情况.获取实验小鼠血清及肿瘤组织.计算各组肿瘤体积,通过CD34免疫组化染色测定各组肿瘤组织的微血管密度(MVD).结果 剂量为0.01、0.1、1 mg/L的RPM对对数生长的H_(22)肝细胞具有细胞毒性,抑制H_(22)小鼠肝癌细胞增殖,培养上清液中的VEGF浓度较对照组显著降低(P<0.05),细胞周期分析显示S期细胞数较其他免疫抑制剂组显著减少(P<0.05).体内实验显示给予1.5 mg/(kg·d)和4.5 mg/(kg·d)的RPM与5 mg/(kg·d) FK506、20 mg/(kg·d) CsA的皮片存活时间相等,而肿瘤体积显著减小(P<0.05).与对照组相比,实验剂量RPM显著降低了荷瘤小鼠血清中的VEGF含量(P<0.05),同时瘤组织内的MVD显著减少(P<0.05).结论 体外实验研究和动物实验证实,RPM具有抗免疫排斥和抗肿瘤增殖的特点,可能在肝移植治疗肝脏恶性肿瘤方面发挥优势.     

基于OpenMP的并行遗传算法求解SAT问题

为了提高SAT （boolean satisfiability） 问题求解效率,在OpenMP （open multi-processing） 编程框架下,将遗传算法与局部搜索算法结合,改进了混合遗传算法中的选择算法,将原有选择操作的时间复杂度降低到O(N)级别. 算法采用OpenMP中的编译制导语句#pragma omp parallel粗粒度并行化驱动混合遗传算法,采用#pragma omp single语句块实现了子种群间个体的同步迁移操作. 与同类算法HCGA （hybrid cloud genetic algorithm）比较分析表明:改进算法HGA （hybrid genetic algorithm）以及并行后的混合遗传算法CGPHGA （coarse-grained parallel hybrid genetic algorithm）在求解成功率和求解效率上都有显著提高,部分问题求解成功率提高达5倍.      

Solving resource-constrained multiple project scheduling problem using timed colored Petri nets

To solve the resource-constrained multiple project scheduling problem (RCMPSP) more effectively, a method based on timed colored Petri net (TCPN) was proposed. In this methodology, firstly a novel mapping mechanism between traditional network diagram such as CPM (critical path method)/PERT (program evaluation and review technique) and TCPN was presented. Then a primary TCPN (PTCPN) for solving RCMPSP was modeled based on the proposed mapping mechanism. Meanwhile, the object PTCPN was used to simulate the multiple projects scheduling and to find the approximately optimal value of RCMPSP. Finally, the performance of the proposed approach for solving RCMPSP was validated by executing a mould manufacturing example.     

国产氨利酮对离体豚鼠乳头肌收缩力和动作电位的影响

本文采用细胞内微电极技术在离体豚鼠心乳头肌上观察了AMR对心室肌收缩力和动作电位的影响.结果表明,25μmol/L的AMR能使心肌收缩力DT增加65.4%、收缩力最大上升速率(dT/dt_(max))增加50.8%、收缩力tPT缩短19.5%;250μmol/L的AMR能使DT增加98.5%、dT/dt_(max)增加91.3%、tPT缩短24.7%。另外,这两个浓度的AMR均能延长心肌细胞动作电位时程和有效不应期。说明AMR确有明显的正性肌力作用。     

Determination of enantiomeric impurity of etomidate by high performance liquid chromatography

Objective To determine enantiomeric impurity of etomidate using high performance liquid chromatography. Methods (R)-etomidate and (S)-etomidate were isopropanol-n-hexane. The flow rate of the mobile phase was 0.5mL/min. The detected wavelength was 242nm. Results (R)-etomidate and (S)-etomidate could be separated completely under these conditions. The precision of (R)-etomidate was 1.57% (n=3). The limit of detection of (R)-etomidate was 4.25ng/mL. The average percentage content of (S)-etomidate was 0.09% in the samples. Conclusion The method was repeatable and sufficiently sensitive to determine the enantiomeric impurity of etomidate. It allows the quantitation of the impurities at the 0.085% (w/w) level relative to etomidate at a concentration of the test solution of 5mg/mL.