Lecture speech recognition using discrete‐mixture HMMs

Most of the state‐of‐the‐art speech recognition systems use continuous‐mixture hidden Markov models (CMHMMs) as acoustic models. On the other hand, it is well known that discrete hidden Markov model (DHMM) systems show poor performance because they are affected by quantization distortion. In this paper, we present an efficient acoustic modeling based on discrete distribution for large‐vocabulary continuous speech recognition (LVCSR). In our previous work, we proposed the maximum a posteriori (MAP) estimation of discrete‐mixture hidden Markov model (DMHMM) parameters and showed that the DMHMM system performed better in noisy conditions than the conventional CMHMM system. However, we conducted the recognition experiments on a read/speech task in which the vocabulary size was only 5k. In addition, the DMHMM was not effective in clean condition in that work. In this paper, we have developed a DMHMM‐based LVCSR system and evaluated the system on a more difficult task under clean condition. In Japan, a large‐scale spontaneous speech database ‘Corpus of Spontaneous Japanese’ has been used as the common evaluation database for spontaneous speech and we used it for our experiments. From the results, it was seen that the DMHMM system showed almost the same performance as the CMHMM system. Moreover, performance improvement could be achieved by a histogram equalization method. Copyright © 2010 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Selective integration of local‐feature detector by boosting for pedestrian detection

An example‐based classification algorithm for pedestrian detection is presented. The classifier integrates component‐based classifiers according to the AdaBoost algorithm. A probability estimate by a kernel‐SVM is used for the outputs of base learners, which are independently trained for local features. The base learners are determined by selecting the optimal local feature according to sample weights determined by the boosting algorithm with cross‐validation. Our method was applied to the MIT CBCL pedestrian image database, and 54 subregions were extracted from each image as local features. The experimental results showed a good classification ratio for unlearned samples. © 2011 Wiley Periodicals, Inc. Electr Eng Jpn, 177(4): 12–22, 2011; Published online in Wiley Online Library ( wileyonlinelibrary.com ). DOI 10.1002/eej.21195     

基于EFPI传感器的GIS局部放电模式识别研究

非本征法布里-帕罗干涉(EFPI)光纤超声传感器可用于气体绝缘全封闭组合电器(GIS)内部的局部放电超声信号检测及模式识别研究,相较于传统的压电式传感器,具有灵敏度高、抗干扰能力强等优点。基于此,文中在充有0.4 MPa SF₆气体的GIS腔体内设置尖端、金属颗粒、悬浮和沿面4种典型的局部放电模型,创新性地利用EFPI传感器对放电超声信号进行检测,提取单次超声脉冲信号波形特征形成特征参数数据库,分别应用概率神经网络(PNN)算法和支持向量机(SVM)算法进行模式识别并比较分析。EFPI传感器检测到的超声信号特征突出,在提取特征参数的基础上,2种模式识别算法均能达到85%以上的平均识别率,且SVM的识别效果要优于PNN。     

An empirical I‐V model of tunneling real‐space transfer transistors for monostable–bistable transition logic element application

We proposed an empirical I‐V model to represent the negative differential resistance (NDR) regime of fabricated tunneling real‐space transfer transistors (TRSTTs). For TRSTTs to have great potential in monostable–bistable transition logic element (MOBILE) design, our model is able to accurately reproduce the NDR regime including gate‐source‐bias‐controlled NDR values and modulated peak to valley drain current ratios. The modeled I‐V curves, tranconductances, and NDRs with multiple gate biases are in good agreement with measured data. The key parameters in the model have clear physical meanings, and the value of these parameters is easy to be extracted directly from the test I‐V curves. The model is used to simulate a practical MOBILE, and excellent agreement between the simulated and measured data was found. Copyright © 2015 John Wiley & Sons, Ltd.     

Fundamental experiments on an air‐cushion‐based force‐sensing device for soft physical human–robot interaction

Physical human–robot interaction (PHRI) is an essential concept for human‐cooperative robots (HCRs). Because the force control for PHRI is based on the contact force between a human and an HCR, it is very important to collect the force data accurately. Furthermore, soft contact during force sensing is also necessary for PHRI for safety and security. As a challenge in force sensing for PHRI using a soft material, we aim to develop in this study a novel force‐sensing device based on an air cushion. First, the physical model of the air cushion is represented by an air cylinder with a piston, which is subject to constraint by nonlinear elasticity. Second, the elastic properties of the air cushion are examined under the assumption that it can be formulated as a function of pneumatic pressure and the contact area with a human. The force applied to the air cushion is estimated on the basis of the physical model and compared to the actual force data in the fundamental experiments. Force is successfully estimated using the proposed physical model, and the advantages of the air‐cushion‐based force‐sensing device (ACFSD) are verified through these fundamental experiments. © 2013 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.