语音识别及端到端技术现状及展望

通过对语音识别技术的发展梳理,简单介绍了语音识别的历史和应用现状,并将传统语音识别的技术和当前的研究进展进行描述.传统语音识别采用基于统计的方法,采用声谱特征,在GMM-HMM混合结构上进行训练和匹配.当前的语音识别模型主要基于深度学习的方法,采用CNN、RNN都可以有效的进行特征提取从而建立声学模型.进一步的研究采用了端到端的技术,避免了多个模型间的误差传导.端到端技术主要有CTC技术和attention技术,最新的模型和方法着重研究了attention技术,并在尝试进行与CTC的融合以达到更好的效果.最后结合作者自身的理解,概括了语音识别当前所面临问题和未来发展方向.     

基于ICA和KPCA人耳识别技术比较

人耳识别技术是一种新的生物识别技术,它以入耳作为识别媒介来进行身份鉴别,但人耳识别的相关理论和方法还不太完善。首先介绍了独立成分分析方法（ICA）和基于核的主成分分析方法（KPCA）的基本原理,然后通过实验得到在分别采用ICA和KPCA方法时,在不同人耳库上的特征提取时间以及采用不同分类器时的入耳识别率。最后通过分析比较实验结果得到基于ICA方法的识别技术和基于KPCA方法的识别技术各自的优点和缺点。     

生物识别新商机

门禁系统采用人脸识别方式、苹果iPhone5s可在home键设置指纹识别等,生物识别技术开始不断深入人们的生活。伴随技术的不断发展,生物识别市场开始步入黄金期,将有越来越多的行业开始应用生物识别技术。     

社交网络中意见领袖的敏感舆论倾向识别

为准确识别意见领袖的敏感舆论倾向,有效把控敏感类舆情的发展,提出基于多任务学习的敏感舆论倾向识别模型(MTL-SA-LSTM)和基于指纹汇聚技术的快速识别模型.以准确识别意见领袖的敏感舆论倾向为目标,兼顾其识别效率.采用指纹汇聚技术关联原始敏感词和变形敏感词,采用语义指纹技术快速识别重复或相似度较高文本的敏感舆论倾向,通过MTL-SA-LSTM模型,对文本中的敏感舆论及舆论倾向两个任务进行识别.对比实验结果表明,该模型具有较高的识别准确率及识别效率.     

BP神经网络在射频识别定位系统中的应用研究

为了提高采用射频识别技术进行定位的精度,针对无源标签射频识别技术及采用BP神经网络对其定位精度的改善进行了研究。首先建立了基于无源标签的射频识别定位系统,之后建立了相应的BP神经网络,并通过实验进行了验证。实验结果表明,在60 cm×50 cm的区域内,通过四角布置四个天线,利用信号强度作为输入信号,采用BP神经网络可以将定位误差控制在2 cm以内,平均欧几里得误差控制在1以内。说明采用BP神经网络可以改善射频识别定位技术的精度。     

大词汇量连续语音识别探讨

文章探讨了中文和英文的大词汇量连续语音识别,讨论了如何设计数据库进行不同识别技术的评测,采用何种评测方法,以及一些代表性的语音识别技术.     

自然手写汉字FS识别法

提出联机识别自然手写汉字的FS识别法。在剖析五笔字根结构和编码原则基础上,对五笔字根作适应性改造,将键盘输入技术与联机识别技术有机融合的一种识别体系。在多库识别体系中首次采用层间分级技术。分析和实验表明：充分考虑了自然手写汉字书写习惯和结构特征,系统有较高稳定性。     

电子商务网站用户访问模式挖掘中的预处理技术

对电子商务网站的用户访问模式挖掘中数据预处理阶段所采用的技术做了全面的研究,主要包括源数据的采集方法以及数据清理、用户识别、会话识别、事务识别、会话子序列生成等所采用的技术。并给出了框架网页过滤、识别搜索引擎Robot产生的访问记录,以及生成用户会话语义序列的方法。     

基于HMM的步态身份识别

随着生物识别悄然兴起,生物识别技术逐渐成为新的身份识别技术。步态识别是生物特征识别技术的一个新兴子领域。文章就是将隐马尔可夫模型(HMM,HiddenMarkovModel)方法运用在步态身份识别中,并进行了其识别性能的研究。该文给出了一个基于HMM的步态身份识别方案,并进行了图像预处理,HMM参数训练和识别的研究,得出了一些有意义的结论。同时在中国科学院自动化研究所提供的CASIA步态数据库上进行了步态身份识别实验,实验结果表明:在侧面视角下采用此方法,具有较好的识别率。     

HMM在汉字识别技术中的应用

研究在汉字识别技术中，采用ＨＭＭ描述汉语文本识别过程、汉语文本识别后处理以及采用 ２Ｄ－ＨＭＭ进行字符识别，以提高系统的识别率和稳定性。     

伦理智能体及其设计:现状和展望

智能体一直是人工智能的主要研究领域之一,任何独立的能够同环境交互并自主决策的实体都可以抽象为智能体.随着人工智能从计算智能到感知智能,再到认知智能的发展,智能体已逐步渗透到无人驾驶、服务机器人、智能家居、智慧医疗、战争武器等人类生活密切相关的领域.这些应用中,智能体与环境、尤其是与人类和社会的交互愈来愈突出,其中的伦理...     

Sixty years of cybernetics

Fundamental aspects of cybernetics, such as goals, problems, methods, tools, brief history, and correlation with other sciences, are considered. Cybernetics in its classical interpretation is the science of information management, communication, and processing. As cybernetics developed, this definition was formalized as the science of methods and processes of information acquisition, storage, processing, analysis, and evaluation, which allows it to apply to decision making in complex control systems. These systems include all engineering, biological, administrative, social, ecological, and economical systems. The main thesis that determined the goals, problems, subject matter, and development of cybernetics as a whole up to the present is the similarity in management and communication processes in machines, living organisms, and both animal and human societies. First of all, these are processes of transfer, storage, and processing of information, i.e., various signals, messages, and data. Any signal and any information may be considered independently from its particular content and destination as a certain choice between two or more values having the known probabilities (selective concept of information). It allows us to treat all processes on the basis of a unified measure and statistical apparatus. The idea of the general theory of control and communication, that is, cybernetics, is based on this hypothesis.     

类ChatGPT大模型发展、应用和前景

生成式人工智能技术自ChatGPT发布以来，不断突破瓶颈，吸引了资本规模投入、多领域革命和政府重点关注。本文首先分析了大模型的发展动态、应用现状和前景，然后从以下3个方面对大模型相关技术进行了简要介绍：1）概述了大模型相关构造技术，包括构造流程、研究现状和优化技术；2）总结了3类当前主流图像—文本的大模型多模态技术；3）介绍了根据评估方式不同而划分的3类大模型评估基准。参数优化与数据集构建是大模型产品普及与技术迭代的核心问题；多模态能力是大模型重要发展方向之一；设立评估基准是比较与约束大模型的关键方法。此外，本文还讨论了现有相关技术面临的挑战与未来可能的发展方向。现阶段的大模型产品已有强大的理解能力和创造能力，在教育、医疗和金融等领域已展现出广阔的应用前景。但同时，它们也存在训练部署困难、专业知识不足和安全隐患等问题。因此，完善参数优化、优质数据集构建、多模态等技术，并建立统一、全面、便捷的评估基准，将成为大模型突破现有局限的关键。     

Information organization in flexible automation systems

Felxible Automation Systems (FAS) are comprised of three major sybsystems: the workstations, material handling, and computer control. Much effort and research have been devoted to the first two and to system control, the loading and scheduling of the system. Little effort has been expended, however, on the information requirements of FASS, although their control is information intensive. Every move of every axis of every device must be coordinated, monitored, and controlled. Information must be stored and transmitted to the needed device at the appropriate time. The status of the system must be continuously surveyed, updated, and verified.Hierarchical systems have been suggested for the control of FMSS, and this has implications for their structure, which in turn influences computing and communication requirements, system performance, reliability, and failure recovery. This paper discusses storage and data flow requirements as a function of the system's manufacturing mission. Two control architectures, centralized and two-level distributed, are considered, and the computing, storage, and communications requirements calculated for each. Analysis of results indicates that there is a break-even point between centralized and decentralized systems that is a function of the manufacturing mission but independent of FAS size and operating environment.     

区块链+物联网在农产品溯源中的应用研究

近年来,农产品安全问题日益严峻,传统的农产品追溯体系面临公信力缺失、监管困境和扩展性问题,农产品质量安全追溯迫在眉睫。随着区块链技术不断发展,其具有的分布式、去中心化、不可篡改、可追溯等特性在改善农产品溯源系统数据安全性、透明性等方面发挥着重要作用,并受到了各个行业的广泛关注。在简述可追溯性和追溯系统概念的基础上,介绍物联网和区块链技术,并探讨了当前国内外学者将区块链和物联网框架融合应用到农产品溯源中的一些相关研究,同时阐述了溯源系统在物联网与区块链结合下带来的安全、可靠、透明等好处及面临可扩展性、效率问题、资源浪费等挑战,最后对当前农产品溯源体系建设提出建议以及未来的研究方向。     

Adaptive filtering under minimum information divergence criterion

Traditional filtering theory is always based on optimization of the expected value of a suitably chosen function of error, such as the minimum mean-square error (MMSE) criterion, the minimum error entropy (MEE) criterion, and so on. None of those criteria could capture all the probabilistic information about the error distribution. In this work, we propose a novel approach to shape the probability density function (PDF) of the errors in adaptive filtering. As the PDF contains all the probabilistic information, the proposed approach can be used to obtain the desired variance or entropy, and is expected to be useful in the complex signal processing and learning systems. In our method, the information divergence between the actual errors and the desired errors is chosen as the cost function, which is estimated by kernel approach. Some important properties of the estimated divergence are presented. Also, for the finite impulse response (FIR) filter, a stochastic gradient algorithm is derived. Finally, simulation examples illustrate the effectiveness of this algorithm in adaptive system training. Recommended by Editorial Board member Naira Hovakimyan under the direction of Editor Jae Weon Choi. This work was supported in part by the National Natural Science Foundation of China under grants 50577037 and 60604010. Badong Chen received the B.S. and M.S. degrees in Control Theory and Engineering from Chongqing University, Chongqing, China, in 1997 and 2003, respectively, and the Ph.D. degree in Computer Science and Technology from Tsinghua University, Beijing China, in 2008. He is currently a Postdoctor of the Institute of Manufacturing Engineering, Department of Precision Instruments and Mechanology, Tsinghua University, Beijing, China. His research interests are in signal processing, adaptive control, and information theoretic aspects of control systems. Yu Zhu received the B.S. of Radio Electronics in 1983 at Beijing Normal University, and the M.S. of Computer Applications in 1993, and the Ph.D. of Mechanical Design and Theory in 2001 at China University of Mining & Technology. He is now a Professor of the Institute of Manufacturing Engineering of Department of Precision and Mechanology of Tsinghua University. His current research interests are parallel machanism and theory, two photon micro-fabrication, ultra-precision motion system and motion control. Jinchun Hu received the Ph.D. in Control Science and Engineering from Nanjing University of Science and Technology, Nanjing, China, in 1998. Since then, he has been a postdoctoral researcher in Nanjing University of Aeronautics and Astronautics in 1999 and Tsinghua University in 2002 respectively. His research interests are in flight control, aerial Robot and intelligent control. Dr. Hu is currently an Associate Professor of the Department of Computer Science and Technology of Tsinghua University, Beijing, China. Zengqi Sun received the B.S. degree from the Department of Automatic Control, Tsinghua University, Beijing, China, in 1966 and the Ph.D. degree in Control Engineering from the Chalmas University of Technology, Sweden, in 1981. He is currently a Professor of the Department of Computer Science and Technology, Tsinghua University, Beijing, China. He is the author or coauthor of more than 100 paper and eight books on control and robotics. His research interests include robotics, intelligent control, fuzzy system, neural networks, and evolutionary computation.     

Fuzzy approach for image retrieval

Current technology allows the acquisition, transmission, storing, and manipulation of large collections of images. Content-based information retrieval is now a widely investigated issue that aims at allowing users of multimedia information systems to retrieve images coherent with a sample image. A way to achieve this goal is the automatic computation of features such as color, texture, and shape and the use of these features as query terms. Feature extraction is a crucial part of any such system. Current methods for feature extraction suffer from two main problems: firstly, many methods do not retain any spatial information, and secondly, the problem of invariance with respect to standard transformation is still unsolved. In this paper, we describe some results of a study on similarity evaluation in image retrieval using shape, texture, and color as content features. Images are retrieved based on similarity of features, where features of the query specification are compared with features of the image database to determine which images match similarly with given features. In this paper, we propose an effective method for image representation which utilizes fuzzy features. The text was submitted by the author in English. Ryszard S. Choraś is Professor of Computer Science in the Department of Telecommunications and EE of University of Technology and Agriculture, Bydgoszcz, Poland. He also holds a courtesy appointment with the Faculty of Mathematics, Technology, and Natural Sciences of Kazimierz Wielki University, Bydgoszcz and the College of Computer Science, Lódz, Poland. His research interests include image signal compression and coding, computer vision, and multimedia data transmission. He received his M.S. degree in Electrical Engineering from Electronics from the Technical University of Wroclaw, Poland in 1973, and his Ph.D. degree in Electronics from Technical University of Wroclaw, Poland, in 1980, and D.Sc. (Habilitation degree) in Computer Science from Warsaw Technical University, Poland, in 1993. Until 1973–1976 he was a member of the research staff at the Institute of Mathematical Machines Silesian Division, Gliwice, working on graphics hardware and human visual perception. In 1976, he joined University of Technology and Agriculture, Bydgoszcz, Poland, first as an Assistant, then as a Professor of Computer Science at the Department of Telecommunications and EE. From 1994 to 1996, he was also Professor of Computer Sciences of the Zielona Góra University, Poland. He has served as the Chairman of the Communication Switching Division and as Chief of the Image Processing and Recognition Group. Until 1996–2002 he was the Vice Rector of University of Technology and Agriculture, Bydgoszcz. Prof. Choraś has an expertise in EU Programs and National Programs, e.g., he was coordinator of EU Program CME-02060, EU Program on Continuous Education and Technology Transfer, and coordinator of national programs in IST and multimedia in e-learning. Prof. Choraś has authored two monographs, and over 130 book chapters, journal articles, and conference papers in the area of image processing. Professor Choraś is a member of the editorial boards of “Machine Vision and Graphics.” He is the editor-in-chief of “Image Processing and Communications Journal.” He has served on numerous conference committees, e.g., Visualization, Imaging, and Image Processing (VIIP), IASTED International Conference on Signal Processing, Pattern Recognition and Applications, International Conference on Computer Vision and Graphics, ICINCO International Conference on Informatics in Control, Automation and Robotics, ICETE International Conference on E-business and Telecommunication Networks, and CORES International Conference on Computer Recognition Systems, and many others. Prof Choraś is a member of the IASTED, WSEAS, various Committees of the Polish Academy of Sciences, TPO. When not working on academic ventures, Professor Choraś likes to relax with activities such as walking, tennis, and swimming.     

车联网隐私保护技术研究

随着汽车智能化、网联化程度的不断加深,车辆、用户及第三方机构之间的数据共享日益成为刚需,由车辆、用户、路边单元等通信实体之间构建的网络车联网应运而生,而车联网的高移动性和网络拓扑多变性使其更容易遭受攻击,进而导致严重的车联网用户隐私泄露问题。如何平衡数据共享和隐私保护之间的关系成为车联网产业发展所面临的一个关键挑战。近年来,学术界针对车联网隐私保护问题进行了深入的研究,并提出了一系列解决方案,然而,目前缺少对这些方案从隐私属性方面进行分析。为此,本文首先从车联网的系统架构、通信场景及标准进行阐述。然后对车联网隐私保护的需求、攻击模型及隐私度量方法进行分析与总结。在此基础上从车联网身份隐私、匿名认证位置隐私和车联网位置服务隐私三个方面出发,介绍了匿名认证、假名变更、同态加密、不经意传输等技术对保护车联网用户隐私起到的重要作用,并讨论了方案的基本原理及代表性实现方法,将方案的隐私性从不可链接性、假名性、匿名性、不可检测性、不可观察性几个方面进行了分析与总结。最后探讨了车联网隐私保护技术当前面临的挑战及进一步研究方向,并提出了去中心化的车辆身份隐私技术以保护车辆身份隐私、自适应假名变更技术以支持匿名认证、满足个性化隐私需求的位置服务隐私保护技术,以期望进一步推动车联网隐私保护技术研究的发展与应用。     

On minimizing errors in 3D reconstruction for stereo camera systems

Active reconstruction of 3D surfaces deals with the control of camer a viewpoints to minimize error and uncertainty in the reconstructed shape of an object. In this paper we develop a mathematical relationship between the setup and focal lengths of a stereo camera system and the corresponding error in 3D reconstruction of a given surface. We explicitly model the noise in the image plane, which can be interpreted as pixel noise or as uncertainty in the localization of corresponding point features. The results can be used to plan sensor positioning, e.g., using information theoretic concepts for optimal sensor data selection. The text was submitted by the authors in English. Stefan Wenhardt, born in 1978, graduated in mathematics at the University of Applied Sciences, Regensburg, Germany, in 2002, with a degree of Dipl.-Math. (FH). Since June 2002, he has been a research staff member at the Chair for Pattern Recognition at the Friedrich-Alexander-University of Erlangen-Nuremberg, Germany. The topics of his research are 3D reconstruction and active vision systems. He is author or coauthor of four publications. Joachim Denzler, born April 16, 1967, received a degree of Diplom-Informatiker, Dr.-Ing. and Habilitation from the University of Erlangen in 1992, 1997, and 2003, respectively. Currently, he holds a position of a full professor for computer science and is head of the computer vision group, Faculty of Mathematics and Informatics, University of Jena. His research interests comprise active computer vision, object recognition and tracking, 3D reconstruction, and plenoptic modeling, as well as computer vision for autonomous systems. He is author and coauthor of over 80 journal papers and technical articles. He is member of the IEEE computer society, DAGM, and GI. For his work on object tracking, plenoptic modeling, and active object recognition and state estimation, he was awarded with the DAGM best paper awards in 1996, 1999, and 2001, respectively. Heinrich Niemann obtained the degree of Dipl.-Ing. in Electrical Engineering and Dr.-Ing. from Technical University Hannover, Germany. He worked at the Fraunhofer Institut fur Informationsverarbeitung in Technik und Biologie, Karlsruhe, and at Fachhochschule Giessen in the department of Electrical Engineering. Since 1975 he has been Professor of Computer Science at the University of Erlangen-Nurnberg, where he was dean of the engineering faculty of the university from 1979–1981. From 1988–2000 he was head of the research group Knowledge Processing at the Bavarian Research Institute for Knowledge-based Systems (FORWISS). Since 1998 he has been the speaker of a special research area entitled Model-based Analysis and Visualization of Complex Scenes and Sensor Data, which is funded by the German Research Foundation (DFG). His fields of research are speech and image understanding and the application of artificial intelligence techniques in these fields. He is on the editorial boards of Signal Processing, Pattern Recognition Letters, Pattern Recognition and Image Analysis, and Journal of Computing and Information Technology. He is the author or coauthor of 7 books and about 400 journal and conference contributions, as well as editor or coeditor of 24 volumes of proceedings and special issues. He is a member of DAGM, ISCA, EURASIP, GI, and IEEE, and a Fellow of IAPR.     

视觉传感机理与数据处理进展

传统视觉感知以RGB光学图像和视频图像为主要数据源,借助计算机视觉的发展取得了巨大成功。然而,传统RGB光学成像也存在着光谱、采样速度、测量精度、可工作条件等方面的限制。近年来,视觉感知的新机理和新数据处理技术的迅速发展,为提升感知和认知能力带来了重大机遇;同时,也具有重要的理论价值和重大应用需求。本文围绕激光扫描、水声声呐成像、新体制动态成像、计算成像、位姿感知等研究方向,综述发展现状、前沿动态、热点问题和发展趋势。当前,在视觉传感研究领域,国内研究机构和团队在数据处理和应用方面取得了显著进展。整体上,国内依然要落后于欧美日等先进国家,尤其是在相关硬件的研制方面。最后,给出了发展趋势与展望,以期为相关研究者提供参考。