Speech Enhancement Using Phase‐Dependent A Priori SNR Estimator in Log‐Mel Spectral Domain

We propose a novel phase‐based method for single‐channel speech enhancement to extract and enhance the desired signals in noisy environments by utilizing the phase information. In the method, a phase‐dependent a priori signal‐to‐noise ratio (SNR) is estimated in the log‐mel spectral domain to utilize both the magnitude and phase information of input speech signals. The phase‐dependent estimator is incorporated into the conventional magnitude‐based decision‐directed approach that recursively computes the a priori SNR from noisy speech. Additionally, we reduce the performance degradation owing to the one‐frame delay of the estimated phase‐dependent a priori SNR by using a minimum mean square error (MMSE)‐based and maximum a posteriori (MAP)‐based estimator. In our speech enhancement experiments, the proposed phase‐dependent a priori SNR estimator is shown to improve the output SNR by 2.6 dB for both the MMSE‐based and MAP‐based estimator cases as compared to a conventional magnitude‐based estimator.     

中国专利转化途径改进初探

基于中国专利申请的实际情况,分析目前专利转化率低的成因.从专利转化途径方面入手,提出通过专利转化政策的调整和相关制度的松绑,以及由国家知识产权局牵头搭建一个专业、权威、规范、统一、监管有力的信息服务平台,为专利权利人和相关企业提供切实服务,为国内知识产权战略的落实提供支撑,促进经济社会的快速、有序、健康发展.     

提高自主创新能力，加速变频器制造业持续快速发展

创新是企业发展不竭的动力，是增强企业竞争力的关键因素。本文就变频器企业在当前激烈的市场矗虿中，如何做好自主创新工作进行了探讨。在简要阐述自主创新的重要意义后，分析了变频器企业目前发展的现状和存在的问题。针对问题提出了变频器企业在开展自主创新时，可供参考的一些基本思路和途径。     

机制主义:人工智能的统一理论

长期以来,学术界曾经分别从智力系统的结构、功能和行为三个不同的侧面分析和模拟人的智力过程,取得了众多成果,形成了人工智能理论的结构主义、功能主义和行为主义三大体系.这些研究也留下许多问题,而且互相缺乏沟通,常常发生"孰优孰劣"的争论.本文试图提出智能生成的"机制主义",希望从智能生成的共性机制入手探讨智能的本质.研究发现:在一般情形下,智能生成机制表现为"信息-知识-智能的转换",由此引出了"知识理论";特别有意义的是,结构主义、功能主义、行为主义三者可以在"机制主义"框架下得到和谐完美的统一.这可能为智能科学技术的发展提供新的理解,带来新的机会.     

Cracking energy estimation of ultra low-k package using novel prediction approach combined with global-local modeling technique

Interfacial crack/delamination, due to the presence of dissimilar material systems, is one of the major concerns of thermo-mechanical reliability for the development of next node technology in integrated circuits (IC) devices. The cracking energy results from many back end of line (BEOL) and packaging processes at various temperature differences is prone to drive the crack advance. To investigate the sensitivity of crack propagation in low-k dielectric materials, a robust estimation of J-integral approach combined with a rectangular path of integral contour is performed using finite element analysis (FEA). By means of the verification of 4-point bending test (4-PBT), excellent agreements are obtained as compared with the experimental data. Moreover, a multiscale modeling technique is proposed to resolve the difficulty of model construction as from bridge device level to packaging level. The sub-modeling procedures developed specifically for the impact prediction of interfacial crack in complicated Cu/low-k interconnects. The analytic results indicate the foregoing methodology is valuable to forecast the physical behavior and reliability of advanced IC devices in the nano scaled size. On the basis of the presented results in this research, an approximated criterion for determining the dimensions of sub-model is suggested and demonstrated as well.     

Mechatronic design of a self-balancing three-dimensional inertia wheel pendulum

In this paper a three-dimensional self-balancing cube is introduced. For stabilization and swing-up, reaction wheels are placed within the cube along with the actuation, electronics and sensors. This allows the cube to swing up from a rest position and to balance around the unstable steady-state positions. We conducted a design assessment to review different design approaches. For the most promising design, we formulated a parameter optimization to calculate the most suitable dimensions for the cube and the inertia elements. These results were used in the mechanical design stage to obtain the optimal design parameters for the cube in an iterative process. To show the effectiveness of the test bench we considered a proportional control law for swing-up and stabilization at the various steady state positions.     

Adaptive–interpolative binarization with stroke preservation for restoration of faint characters in degraded documents

A novel technique for binarization with stroke preservation of faint characters in degraded documents is proposed. It works in a multi-scale framework with an adaptive–interpolative thresholding technique. Instead of computing a global threshold value, it computes the local threshold values for a small set of grid points by observing the intensity pattern of the pixels lying in the concerned grid cells. Estimated thresholds are used, in turn, to compute the threshold values of all the remaining pixels using a fast-yet-efficient interpolation procedure. To handle noises in degraded images, this grid-based adaptive thresholding is applied in successively reducing scales to obtain the near-optimal binarization as a set of connected components. After a post-processing meant for stroke preservation with these connected components, we get the final output. Exhaustive experimentation and comparison with other existing methods have been successfully carried out with benchmark datasets and also with our own datasets.     

A COMPUTER-AIDED UNIFIED APPROACH TO MODELING PWM AND RESONANT CONVERTERS

This letter puts forward a method of modeling for the steady-state and small signal dynamic analysis on PWM, quasi-resonant and series/（parallel） resonant switching converters based on pulse-waveform integral approach. As an example, PWM and quasi-resonant converters are used to discuss the principle of the approach. The results are compared with those in the relative literatures. Computer aided analysis are made to confirm the correctness.     

Value drivers of blockchain technology: A case study of blockchain-enabled online community

There is growing recognition that blockchain technology has significant potential to alter how organizations and people work and communicate. However, theoretical guidance concerning how organizations leverage blockchain technology to enhance value creation for users is still limited. Grounded in the socio-technical perspective and leveraging the rich data obtained from case analyses of blockchain-enabled online communities, this paper develops a theoretical model to identify the core value drivers that blockchain enables for online communities. The core value drivers include: a reputation-value system, data ownership mechanisms, and verification & tracking mechanisms. Our findings suggest that these three value drivers enhance value creation of online communities by motivating participation and protecting contributions.     

基于线性迭代的DFT调制滤波器组的设计算法

该文研究了单原型的DFT调制滤波器组的设计方法。在该方法中,滤波器组的设计问题被归结为一个无约束的优化问题,其目标函数为滤波器组的传递失真、混叠失真和原型滤波器的阻带能量的加权和。结合线性化方法,原型滤波器通过迭代求解。在单步迭代中,原型滤波器的系数通过解析式求解得到。仿真表明,与传统的设计算法相比,本文方法设计所得的DFT调制滤波器组重构误差减小了约40dB,阻带衰减提高了约2dB。并且新算法的计算复杂度明显低于传统算法。