检索结果-内蒙古大学图书馆

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Felix Weninger Jean-Louis Durrieu Florian Eyben Gael Richard Bjorn Schuller Institute for Human-Machine Communication Technische Universitat Munchen D-80290 Munchen Signal Processing Laboratory Ecole Polytechnique Federale de Lausanne CH-1015 Lausanne Institut Telecom Telecom ParisTech LTCI-CNRS F-75014 Paris

ISBN: (纸本)9781457705380

We present a novel and unique combination of algorithms to detect the gender of the leading vocalist in recorded popular music. Building on our previous successful approach that enhanced the harmonic parts by means of Non-Negative Matrix Factorization (NMF) for increased accuracy, we integrate on the one hand a new source separation algorithm specifically tailored to extracting the leading voice from monaural recordings. On the other hand, we introduce Bidirectional Long Short-Term Memory Recurrent Neural Networks (BLSTM-RNNs) as context-sensitive classifiers for this scenario, which have lately led to great success in Music Information Retrieval tasks. Through a combination of leading voice separation and BLSTM networks, as opposed to a baseline approach using Hidden Naive Bayes on the original recordings, the accuracy of simultaneous detection of vocal presence and vocalist gender on beat level is improved by up to 10 percent absolute. Furthermore, using this technique we achieve 91.6 percent accuracy in determining the gender of the predominant vocalist on song level, which is 4 percent absolute above our previous best result.

关键词： Long Short-Term Memory Non-Negative Matrix Factorization Music Information Retrieval Gender composite algorithm Voice Recordings Music information retrieval Robustness separation Source separation

来源：评论

学校读者我要写书评

暂无评论

Description of the JHU System Combination Scheme for WMT 2011 11

Description of the JHU System Combination Scheme for WMT 201...

引用

Workshop on statistical machine translation

作者： Daguang Xu Yuan Cao Damianos Karakos Johns Hopkins University Baltimore

ISBN: (纸本)9781937284121

This paper describes the JHU system combination scheme used in WMT-11. The JHU system combination is based on confusion network alignment, and inherited the framework developed by (Karakos et al., 2008). We improved our core system combination algorithm by making use of TER-plus, which was originally designed for string alignment, for alignment of confusion networks. Experimental results on French-English, German-English, Czech-English and Spanish-English combination tasks show significant improvements on BLEU and TER by up to 2 points on average, compared to the best individual system output, and improvements compared with the results produced by ITG which we used in WMT-10.

关键词： Belgo-Luxembourg Economic Union ITG TECR gene Alignment Confusion composite algorithm output(systems) individual system Frameworks

来源：评论

学校读者我要写书评

暂无评论

Infant cry recognition based on feature extraction

Infant cry recognition based on feature extraction

引用

International Conference on Information Networking and Automation (ICINA)

作者： Feng-Ling Yuan Xiang-Nian Huang School of Mathematics and Computer Engineering Xihua University Chengdu China

ISBN: (纸本)9781424481040

The choice of feature parameters is very crucial for the system of infant cry recognition. This paper describes the principle and method of mel cepstrum coefficients which is used commonly and adopts a combination algorithm of mel frequency cepstrum coefficients, cepstrum mean subtraction, inverted mel frequency cepstral coefficient, which aim to enhance recognition accuracy in the infant cry recognition system. Results show that the recognition rate of improved algorithm increases compared to the classical algorithm in the same environment.

关键词： feature extraction infa-nt recognition mel-frequency cepstrum coefficients (mfcc) Feature extraction crying infant composite algorithm Cepstrum Recognition (Psychology) Mel frequency cepstral coefficient

来源：评论

学校读者我要写书评

暂无评论

CLUSTERING-DRIVEN WIRELESS DATA BROADCASTING

引用

IEEE WIRELESS COMMUNICATIONS 2009年第6期16卷 80-87页

作者： Liaskos, Christos K. Petridou, Sophia G. Papadimitriou, Georgios I. Nicopolitidis, Petros Pomportsis, Andreas S. Obaidat, Mohammad S. Aristotle Univ Thessaloniki Dept Informat Thessaloniki Greece Monmouth Univ Grad Program W Long Branch NJ USA Monmouth Univ Dept Comp Sci W Long Branch NJ USA

The performance of a push-based system relies heavily on the proper scheduling of the broadcast data. To this end, the Broadcast Disks method is most commonly employed. It defines a procedure consisting of four separate algorithms: one to provide and handle the clients' feedback, another to group the data objects into disks, a third one to define their spinning velocities, and finally a Broadcast Sequence constructor algorithm. In this article we introduce and evaluate Clustering-Driven Wireless Data Broadcasting (CWDB), a complete instantiation of the Broadcast Disks method. The proposed CWDB procedure addresses the major omissions of preceding schemes, including the total lack of feedback mechanisms, extremely limited variety of spinning velocity definition algorithms, and no thorough and realistic testing of complete combinations of algorithms under various client configurations. A new efficient clustering-driven data grouping algorithm is also introduced. Extensive simulation results are presented, which indicate that the proposed CWDB is absolutely dominant over other classical methods in the vast majority of the test cases and led to very important conclusions that were previously overlooked.

关键词： Simulation results Spinning composite algorithm feedback mechanism algorithms wireless data Spin CASING INSPECTION broadcasting

来源：评论

学校读者我要写书评

暂无评论

Cross sections for ionization of K, L and M shells of atoms by impact of electrons and positrons with energies up to 1 GeV: Analytical formulas

引用

ATOMIC DATA AND NUCLEAR DATA TABLES 2009年第6期95卷 871-909页

作者： Bote, David Salvat, Francesc Jablonski, Aleksander Powell, Cedric J. Univ Barcelona Fac Fis ECM & ICC E-08028 Barcelona Spain Polish Acad Sci Inst Phys Chem PL-01224 Warsaw Poland NIST Surface & Microanal Sci Div Gaithersburg MD 20899 USA

Analytical formulas are presented for the easy calculation of cross sections for ionization of K. L and M shells of neutral atoms by impact of electrons and positrons with kinetic energies up to 1 GeV. Each formula contains a number of parameters that are characteristic of the element, the active electron shell and the projectile particle. The values of these parameters were determined by fitting the cross section values in an extensive database that was calculated recently by means of a composite algorithm that combines the distorted-wave and plane-wave Born approximations. Tables of parameter Values are given for all elements. from hydrogen (Z = 1) to einsteinium (Z = 99). The proposed analytical expressions yield ionization cross sections that agree with those in the numerical database to within about 1%, except for projectiles with near-threshold energies. (C) 2009 Elsevier Inc. All rights reserved.

关键词： ionization positron calculation Electrons composite algorithm POSITRONS Atoms Neutral atoms formula IONIZATION CROSS SECTIONS M SHELL Cross section ionization

来源：评论

学校读者我要写书评

暂无评论

The mid-trimester genetic sonogram

引用

SEMINARS IN PERINATOLOGY 2005年第4期29卷 209-214页

作者： Bahado-Singh, RO Oz, UA Mendilcioglu, I Mahoney, MJ Univ Cincinnati Coll Med Dept Obstet & Gynecol Cincinnati OH 45267 USA Univ Mersin Fac Med Dept Obstet & Gynecol Zeytinlibahce Mersin Turkey Akdeniz Univ Sch Med Dept Med Biol & Genet Antalya Turkey Yale Univ Sch Med Dept Med Biol & Genet New Haven CT USA Yale Univ Sch Med Dept Genet New Haven CT 06510 USA

The genetic sonogram is a composite algorithm combining multiple individual markers to increase Down syndrome risk prediction. Transformation of sonographic information into a standard mathematical format represented an early challenge that has now been surmounted. Using increasingly sophisticated mathematical techniques, individual patient risk can be estimated. High diagnostic accuracy comparable to standard mid-trimester serum algorithms has been reported. Most recently, a few studies have reported the ability to combine serum and biochemical markers to achieve diagnostic accuracy comparable to first-trimester screen. Even fewer studies have reported combinations of ultrasound and maternal urine markers. While it is clear that consistently high sensitivity and specificity for Down syndrome can be achieved, almost all the studies are based on high-risk groups. Studies in low-risk populations have suffered from lack of standardization. The relevance of genetic sonogram in a low-risk population thus remains to be proven. The most significant challenge, however, remains the development of uniform and reproducible sonographic and measurement standards. This is likely to be the most important factor in optimizing the accuracy of the mid-trimester genetic sonogram.

关键词： Genetic Ultrasonography Risk diagnostic accuracy 超声检查危险 composite algorithm diagnosis quality/standard Sonogram Ultrasonography measurement standards Population at Risk

来源：评论

学校读者我要写书评

暂无评论

Development of the IM147: An alternative inspection/maintenance mass-emission transient test to address vehicle preconditioning concerns

引用

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 2004年第3期54卷 269-285页

作者： Joy, RW Heirigs, PL Torgerson, GD St Denis, M Austin, TC Gordon, J Tefft, B Lindner, J Sierra Res Inc Sacramento CA USA Gordon Darby Inc Louisville KY USA US EPA Ann Arbor MI USA

A series of studies was performed to develop an alternative to the U.S. Environmental Protection Agency's gold standard IM240 mass-based emission test. The new IM147 test was based on the second phase of the IM240 that consists of 147 sec of transient vehicle operation. Paired IM240/IM147 tests were conducted on vehicles ranging from 1981 to 1996 to determine IM147 cutpoints and excess emissions were identified. Additionally, an optimized test procedure was developed that combined possible triplicate IM147s with improved drive trace quality control, fast-pass, and retest methods. The optimized procedure was found to provide improved vehicle preconditioning with a relatively minor decrease in excess emissions identification. Resulting identification rates ranged from 96 to 100% for hydrocarbons (HC), 93-100% for CO, and 93-100% for NO., depending on cutpoint selection, while false failures caused by lack of vehicle preconditioning were reduced to essentially zero. Significant vehicle throughput improvements were achieved through the development of software algorithms involving modal fast-pass and retest procedures. Modal drive trace variation limits also were developed to improve test accuracy. The combination of the algorithms reduced average IM147 test times by nearly 60%.

关键词： composite algorithm Vehicle operations CONCERNS second phase Hydrocarbons vehicle Software algorithms Testing Procedures Pretreatment

来源：评论

学校读者我要写书评

暂无评论

The Use of Classifiers in Sequential Inference 00

The Use of Classifiers in Sequential Inference

引用

Annual Conference on Neural Information Processing Systems

作者： Dan Roth Vasin Punyakanok Department of Computer Science University of Illinois at Urbana-Champaign Urbana IL 61801

ISBN: (纸本)0262122413

We study the problem of combining the outcomes of several different classifiers in a way that provides a coherent inference that satisfies some constraints. In particular, we develop two general approaches for an important subproblem - identifying phrase structure. The first is a Marko-vian approach that extends standard HMMs to allow the use of a rich observation structure and of general classifiers to model state-observation dependencies. The second is an extension of constraint satisfaction formalisms. We develop efficient combination algorithms under both models and study them experimentally in the context of shallow parsing.

关键词： Classifiers composite algorithm Sequential constraint satisfaction problems Parsing

来源：评论

学校读者我要写书评

暂无评论

ON RECURRENT AND RECURSIVE INTERCONNECTION PATTERNS

引用

INFORMATION PROCESSING LETTERS 1986年第6期22卷 285-289页

作者： PARBERRY, I Department of Computer Science College of Science The Pennsylvania State University 333 Whitmore Laboratory University Park PA 16802 U.S.A.

Many trivalent graphs, in particular variants of the cube-connected cycles and shuffle-exchange, have become popular as interconnection patterns for synchronous parallel computers. Highly-structured interconnection patterns are considered that allow large parallel machines to be constructed from isomorphic copies of smaller ones, plus possibly a few extra processors. If only a small number of extra processors are added, the interconnection pattern is called recurrent. If no extra processors are added, it is called recursive. It is demonstrated that a constant-degree recursive interconnection pattern is, in a sense, not as versatile as the cube-connected cycles or shuffle-exchange, and a trivalent recurrent interconnection pattern that is as versatile is presented.

关键词： Parallel computation interconnection pattern composite algorithm cube-connected cycles

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：