基于本征音子说话人子空间的说话人自适应算法

本征音子说话人自适应算法在自适应数据量充足时可以取得很好的自适应效果,但在自适应数据量不足时会出现严重的过拟合现象。为此该文提出一种基于本征音子说话人子空间的说话人自适应算法来克服这一问题。首先给出基于隐马尔可夫模型-高斯混合模型(HMM-GMM)的语音识别系统中本征音子说话人自适应的基本原理。其次通过引入说话人子空间对不同说话人的本征音子矩阵间的相关性信息进行建模;然后通过估计说话人相关坐标矢量得到一种新的本征音子说话人子空间自适应算法。最后将本征音子说话人子空间自适应算法与传统说话人子空间自适应算法进行了对比。基于微软语料库的汉语连续语音识别实验表明,与本征音子说话人自适应算法相比,该算法在自适应数据量极少时能大幅提升性能,较好地克服过拟合现象。与本征音自适应算法相比,该算法以较小的性能牺牲代价获得了更低的空间复杂度而更具实用性。     

Toward Robust Speech Recognition and Understanding

The principal cause of speech recognition errors is a mismatch between trained acoustic/language models and input speech due to the limited amount of training data in comparison with the vast variation of speech. It is crucial to establish methods that are robust against voice variation due to individuality, the physical and psychological condition of the speaker, telephone sets, microphones, network characteristics, additive background noise, speaking styles, and other aspects. This paper overviews robust architecture and modeling techniques for speech recognition and understanding. The topics include acoustic and language modeling for spontaneous speech recognition, unsupervised adaptation of acoustic and language models, robust architecture for spoken dialogue systems, multi-modal speech recognition, and speech summarization. This paper also discusses the most important research problems to be solved in order to achieve ultimate robust speech recognition and understanding systems. Dr. Sadaoki Furui is currently a Professor at Tokyo Institute of Technology, Department of Computer Science. He is engaged in a wide range of research on speech analysis, speech recognition, speaker recognition, speech synthesis, and multimodal human-computer interaction and has authored or coauthored over 450 published articles. From 1978 to 1979, he served on the staff of the Acoustics Research Department of Bell Laboratories, Murray Hill, New Jersey, as a visiting researcher working on speaker verification. He is a Fellow of the IEEE, the Acoustical Society of America and the Institute of Electronics, Information and Communication Engineers of Japan (IEICE). He was President of the Acoustical Society of Japan (ASJ) from 2001 to 2003 and the Permanent Council for International Conferences on Spoken Language Processing (PC-ICSLP) from 2000 to 2004. He is currently President of the International Speech Communication Association (ISCA). He was a Board of Governor of the IEEE Signal Processing Society from 2001 to 2003. He has served on the IEEE Technical Committees on Speech and MMSP and on numerous IEEE conference organizing committees. He has served as Editor-in-Chief of both Journal of Speech Communication and the Transaction of the IEICE. He is an Editorial Board member of Speech Communication, the Journal of Computer Speech and Language, and the Journal of Digital Signal Processing. He has received the Yonezawa Prize and the Paper Awards from the IEICE (1975, 88, 93, 2003), and the Sato Paper Award from the ASJ (1985, 87). He has received the Senior Award from the IEEE ASSP Society (1989) and the Achievement Award from the Minister of Science and Technology, Japan (1989). He has received the Technical Achievement Award and the Book Award from the IEICE (2003, 1990). He has also received the Mira Paul Memorial Award from the AFECT, India (2001). In 1993 he served as an IEEE SPS Distinguished Lecturer. He is the author of “Digital Speech Processing, Synthesis, and Recognition” (Marcel Dekker, 1989, revised, 2000) in English, “Digital Speech Processing” (Tokai University Press, 1985) in Japanese, “Acoustics and Speech Processing” (Kindai-Kagaku-Sha, 1992) in Japanese, and “Speech Information Processing” (Morikita, 1998) in Japanese. He edited “Advances in Speech Signal Processing” (Marcel Dekker, 1992) jointly with Dr. M.M. Sondhi. He has translated into Japanese “Fundamentals of Speech Recognition,” authored by Drs. L.R. Rabiner and B.-H. Juang (NTT Advanced Technology, 1995) and “Vector Quantization and Signal Compression,” authored by Drs. A. Gersho and R. M. Gray (Corona-sha, 1998).     

软交换下用于协议适配的统一消息的设计

软交换是一代网络的核心,协议适配是软交换的重要功能,下一代网络的重要目标之一就是实现网络融合。在分析软交换体系结构的基础上,提出软交换中协议适配的方法,设计了一个适用于目前网络上存在的所有呼叫控制协议的统一格式消息集,并描述了消息的结构,讨论了统一格式消息与SIP和ISUP之间的映射流程。以该方法为基础所实现的软交换具有较高的性能和可扩展性。     

无线Mesh网络速率自适应算法研究

自适应速率选择是影响无线Mesh网性能的关键问题。提出了一种基于动态解调门限的速率自适应选择算法,该算法通过统计传输速率进行速率选择,能够针对不同信道环境对每种调制速率所需最小信噪比门限进行自动调整。同时,采用跨层设计思想,采用路由协议广播报文作为MESH设备间传输信噪比信息的载体,从而能够根据信噪比和传输成功率信息得到信道环境下的最佳发送速率。实际测试结果表明,所提算法能够在较短时间内选择出最佳速率,而且具有较低的抖动和稳定的吞吐量。     

Defect Engineering in Ambipolar Layered Materials for Mode-Regulable Nociceptor

Defect engineering represents a significant approach for atomically thick 2D semiconductor material development to explore the unique material properties and functions. Doping-induced conversion of conductive polarity is particularly beneficial for optimizing the integration of layered electronics. Here, controllable doping behavior in palladium diselenide (PdSe₂) transistor is demonstrated by manipulating its adatom-vacancy groups. The underlying mechanisms, which originate from reversible adsorption/desorption of oxygen clusters near selenide vacancy defects, are investigated systematically via their dynamic charge transfer characteristics and scanning tunneling microscope analysis. The modulated doping effect allows the PdSe₂ transistor to emulate the essential characteristics of photo nociceptor on a device level, including firing signal threshold and sensitization. Interestingly, electrostatic gating, acting as a neuromodulator, can regulate the adaptive modes in nociceptor to improve its adaptability and perceptibility to handle different danger levels. An integrated artificial nociceptor array is also designed to execute unique image processing functions, which suggests a new perspective for extension of the promise of defect engineered 2D electronics in simplified sensory systems toward use in advanced humanoid robots and artificial visual sensors.     

无线Ad Hoc网络MAC层速率自适应技术研究

目前的IEEE802.11标准虽然在物理层提供了对多速率的支持,但是在MAC（媒体接入控制,Media Access Control）层却没有规定速率自适应的方法。研究了无线Ad Hoc网络（自组织网）中的速率自适应技术,针对RBAR协议存在的不足提出了一种改进的MAC层速率自适应协议,该协议实现简单,与标准的兼容性好,对信道的变化适应能力更强。仿真结果表明,该协议比RBAR协议具有更好的性能。     

Bandwidth Adaptation Algorithms for Adaptive Multimedia Services in Mobile Cellular Networks

The fluctuation of available link bandwidth in mobilecellular networks motivates the study of adaptive multimediaservices, where the bandwidth of an ongoing multimedia call can bedynamically adjusted. We analyze the diverse objectives of theadaptive multimedia framework and propose two bandwidth adaptationalgorithms (BAAs) that can satisfy these objectives. The firstalgorithm, BAA-RA, takes into consideration revenue and``anti-adaptation' where anti-adaptation means that a user feelsuncomfortable whenever the bandwidth of the user's call ischanged. This algorithm achieves near-optimal total revenue withmuch less complexity compared to an optimal BAA. The secondalgorithm, BAA-RF, considers revenue and fairness, and aims at themaximum revenue generation while satisfying the fairnessconstraint defined herein. Comprehensive simulation experimentsshow that the difference of the total revenue of BAA-RA and thatof an optimal BAA is negligible. Also, numerical results revealthat there is a conflicting relationship between anti-adaptationand fairness.     

Temporal information based GoP adaptation for linear video delivery schemes

An original wireless video transmission scheme called SoftCast has been recently proposed to deal with the issues encountered in conventional wireless video broadcasting systems (e.g. cliff effect). In this paper, we evaluate and optimize the performance of the SoftCast scheme according to the transmitted video content. Precisely, an adaptive coding mechanism based on GoP-size adaptation, which takes into account the temporal information fluctuations of the video, is proposed. This extension denoted Adaptive GoP-size mechanism based on Content and Cut detection for SoftCast (AGCC-SoftCast) significantly improves the performance of the SoftCast scheme. It consists in modifying the GoP-size according to the shot changes and the spatio-temporal characteristics of the transmitted video. When hardware capacities, such as buffer or processor performance are limited, an alternative method based only on the shot changes detection (AGCut-SoftCast) is also proposed. Improvements up to 16 dB for the PSNR and up to 0.55 for the SSIM are observed with the proposed solutions at the cut boundaries. In addition, temporal visual quality fluctuations are reduced under 1dB in average, showing the effectiveness of the proposed methods.     

MIMO-OFDM系统速率自适应传输方案

在采用TDD工作方式的MIMO-OFDM实际通信系统中,由于收发两端环境的不同导致干扰对称的假设不再满足。为了弥补这一影响,准确地确定系统的最佳传输参数,提出了一种基于事件驱动机制的闭环自适应传输方案。该方案通过在收发双方建立一条闭环控制信息传输链路,实现了迅速准确地确定最佳传输参数的目的。实验仿真结果表明,当在MIMO-OFDM系统中采用该方案时,可以有效提高系统的吞吐量,减小系统的中断概率。     

Spectral Efficiency Evaluation for MRC Diversity Schemes Over Generalized Rician Fading Channels

In this paper, closed-form expressions for the capacities per unit bandwidth for generalized Rician fading channels are derived for power and rate adaptation, constant transmit power, channel inversion with fixed rate, and truncated channel inversion adaptation policies. The closed-form solutions are derived for the single antenna reception (without diversity combining) and maximal-ratio combining (MRC) diversity cases. Truncated channel inversion adaptation policy is the best policy for the single antenna reception case, while the channel inversion with fixed rate policy is the best policy for the MRC diversity case. Constant transmit power policy provides the lowest spectral efficiency as compared to the other policies with and without diversity.