视觉和语音加工难度对汉语识字困难儿童形音特征捆绑的影响

研究选取汉语识字困难儿童与普通儿童各28名,通过三种图形和语音信息不同加工难度条件,考察不同视觉和语音加工难度对汉语识字困难儿童形音特征捆绑成绩的影响.实验结果显示:(1)在视觉-语音信息加工难度低的时候,汉语识字困难儿童形音特征捆绑正确率显著低于普通儿童,说明其形音特征捆绑落后于普通儿童.(2)在视觉信息加工难度增加...     

量子高密度编码原理及其MATLAB仿真

由于没有实际操纵量子比特的条件,MATLAB以其简便的矩阵操纵手段和强大的科学计算功能,成为通信工程人员和在校学生学习量子通信基本原理的良好工具。给出了量子比特及其测定、量子序列、贝尔态基和常用量子逻辑门的MATLAB语言描述。分析了量子高密度编码原理,给出了MATLAB仿真程序。     

基于改进哈夫曼编码的数据压缩方法研究

作为一种无损压缩编码方法,哈夫曼编码在数据压缩中具有重要的应用。经典的哈夫曼编码是在构造哈夫曼的基础上自下而上进行的,通过分析哈夫曼算法的思想,给出了一种改进的哈夫曼数据压缩算法。该算法利用队列结构,从哈夫曼的根节点出发,向叶子节点进行编码,在编码过程中仅将哈夫曼树的每个叶子节点进行一次扫描便可以得到各个叶子节点的哈夫曼编码。实验表明,改进算法不仅压缩率高于以往算法,而且保证了最终生成的压缩文件的安全性。     

基于小波变换的图像压缩算法设计

本文设计并实现了一个有效的图像压缩算法。选取了Daubechies-5小波基,通过对小波系数的特点的研究,提出了将嵌入式零树编码和改进的游程编码相结合的方法,实现了小波系数的有效编码,实验表明,该算法取得较好的压缩效果。     

基于CNN与Bi-LSTM的唇语识别研究

针对唇语识别过程中唇部特征提取和时序关系存在的问题,提出一种卷积神经网络（CNN）和双向长短时记忆网络（Bi-LSTM）相结合的深度学习模型。利用CNN学习唇部特征,并将学习到的唇部特征送入Bi-LSTM进行时序编码,通过Softmax进行分类。建立NUMBER DATASET和PHRACE DATASET两个大型汉语数据集以解决汉语唇语数据缺失问题。将该模型与传统的唇语识别方法在两个数据集上进行实验对比,发现在NUMBER DATASET上识别准确率为81.3%,比传统方法提高了8.1%,在PHRACE DATASET上识别准确率为83.5%,比传统方法提高了9%。实验结果表明该模型能有效提高唇语识别的准确率。     

Optimum weighting-coefficient-pair in inter-carrier interference self-cancellation scheme of OFDM system

As one of the most important components of the wideband wireless access technique, orthogonal frequency division multiplexing （OFDM） has a high usage rate of spectrum and combats inter-symbol interference （ISI） in multi-path fading channel. However, when there are frequency offsets during the signal transmission, the inter-carrier interference （ICI） is introduced, which significantly degrades the performance. The existing ICI self-cancellation schemes such as PCC-OFDM are not optimum to minimize the interference considering both noise and ICI. In this paper, a new metric named S1NR （signal-to-interference- and-noise ratio） is proposed. We discuss the optimization issue when a constant frequency offset exists and in time-varying channels. The optimum weighting-coefficient-pair （OWCP） is obtained, which maximizes SINR theoretically through the alternant iteration algorithm. Simulations show that the performance of OWCP-OFDM is better than that of PCC-OFDM, especially when the frequency offset is large. Although the ICI self-cancellation scheme suffers bandwidth inefficiency, from the simulation results we can also see that the performance of OWCP-OFDM is much better than that of the standard OFDM systems with the same bandwidth efficiency when a frequency offset exists. Moreover, since the redundant modulation provides the capability to suppress ICI as well as a receiving SNR gain, it can be considered as exchanging the bandwidth for SNR.     

Power duality for multi-antenna OFDM system in broadcast channel with user scheduling

This paper deals with design and analysis of user scheduling and power allocation for multi-antenna OFDM systems with DPC, ZF-DPC, ZF-BF and TDMA transmit strategies. We consider the general multi-user downlink scheduling problem and power minimization with multi-user rate constraints. According to the channel state, it is shown that there is a power optimal policy which selects a subset of users in each scheduling interval. We present user selection algorithms for DPC, ZF-DPC, ZF-BF and TDMA for multi-antenna OFDM system in broadcast channels, and we also present the practical water-filling solution in this paper. By the selected users with the consideration of fairness, we derive the power optimization algorithm with multi-user rate constraints. We also analyze the power duality of uplink-downlink for the transmit strategies of DPC, ZF-DPC and ZF-BF. Simulation results show that the present user-scheduling algorithm and power minimization algorithm can achieve good power performance, and that the scheduling algorithm can guarantee fairness.     

获取五笔字型编码表通用程序

本文模拟ＩＢＭ机的ｉｎｔ９功能，自动操作键盘输入，无需深入各种汉字系统的五笔字型转换表得到完整的输入法编码表，并生成格式符合Ｗｉｎｄｏｗｓ转换码表文件要求的文本文件。     

Visual Basic 6.0编程中错误的防范与处理

掌握Visual Basic 6.0编程中错误的防范与处理方法,对于提高编程能力,建立完善的编程思路,实现程序的最优化,具有重要的现实意义与指导意义.     

基于DWT的遗传矢量量化图像压缩编码研究

将遗传算法应用于小波变换与矢量量化技术相结合的图像压缩编码中,提出了基于DWT和遗传算法的矢量量化图象压缩编码算法.研究表明,在相同压缩比下,这种算法较其它算法能更好地改善图像的质量.