检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Paul, Dipjyoti Shifas, Muhammed P.V. Pantazis, Yannis Stylianou, Yannis Computer Science Department University of Crete Greece Inst. of Applied and Computational Mathematics Foundation for Research and Technology Hellas Greece

The increased adoption of digital assistants makes text-to-speech (TTS) synthesis systems an indispensable feature of modern mobile devices. It is hence desirable to build a system capable of generating highly intelligible speech in the presence of noise. Past studies have investigated style conversion in TTS synthesis, yet degraded synthesized quality often leads to worse intelligibility. To overcome such limitations, we proposed a novel transfer learning approach using Tacotron and WaveRNN based TTS synthesis. The proposed speech system exploits two modification strategies: (a) Lombard speaking style data and (b) Spectral Shaping and Dynamic Range Compression (SSDRC) which has been shown to provide high intelligibility gains by redistributing the signal energy on the time-frequency domain. We refer to this extension as Lombard-SSDRC TTS system. Intelligibility enhancement as quantified by the Intelligibility in Bits (SIIBGauss) measure shows that the proposed Lombard-SSDRC TTS system shows significant relative improvement between 110% and 130% in speech-shaped noise (SSN), and 47% to 140% in competing-speaker noise (CSN) against.the state-of-the-art TTS approach. Additional subjective evaluation shows that Lombard-SSDRC TTS successfully increases the speech intelligibility with relative improvement of 455% for SSN and 104% for CSN in median keyword correction rate compared to the baseline TTS method. Copyright © 2020, The Authors. All rights reserved.

关键词： Speech intelligibility

来源：评论

学校读者我要写书评

暂无评论

Classification-based improvement of application robustness and quality of service in probabilistic computer systems

引用

25th International Conference on Architecture of Computing Systems, ARCS 2012

作者： Heinig, Andreas Mooney, Vincent J. Schmoll, Florian Marwedel, Peter Palem, Krishna Engel, Michael Computer Science 12 TU Dortmund Germany Inst. Sustainable and Applied Infodynamics NTU Singapore Singapore School of ECE Georgia Institute of Technology Atlanta GA United States Department of Computer Science Rice University Houston TX United States

ISBN: (纸本)9783642282928

Future semiconductors no longer guarantee permanent deterministic operation. They are expected to show probabilistic behavior due to lowered voltages and shrinking structures. Compared to radiation-induced errors, probabilistic systems face increased error frequencies leading to unexpected bit-flips. Approaches like probabilistic CMOS provide methods to control error distributions which reduce the error probability in more significant bits. However, inst.uctions handling control flow or pointers still expect deterministic operation, thus requiring a classification to identify these inst.uctions. We apply our transient error classification to probabilistic circuits using differing voltage distributions. Static analysis ensures that probabilistic effects only affect unreliable operations which accept a certain level of impreciseness, and that errors in probabilistic components will never propagate to critical operations. To evaluate, we analyze robustness and quality-of-service of an H.264 video decoder. Using classification results, we map unreliable arithmetic operations onto probabilistic components of a simulated ARM-based architecture, while the remaining operations use deterministic components. © 2012 Springer-Verlag.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

SOFSEM 2001: Theory and Practice of Informatics 1

引用

丛书名： Lecture Notes in computer science

1000年

作者： Leszek Pacholski Peter Ružička

ISBN: (数字)9783540456278

ISBN: (纸本)9783540429128

SOFSEM 2001, the International Conference on Current Trends in Theory and Practice of Informatics, was held on November 24 – December 1, 2001 in the ? well-known spa Pie?stany, Slovak Republic. This was the 28th annual conference in the SOFSEM series organized either in the Slovak or the Czech Republic. SOFSEM has a well-established tradition. Currently it is a broad, multid- ciplinary conference, devoted to the theory and practice of software systems. Its aim is to foster cooperation among professionals from academia and industry working in various areas of informatics. The scienti?c program of SOFSEM consists of invited talks, which determine the topics of the conference, and short contributed talks presenting original - sults. The topics of the invited talks are chosen so as to cover the whole range from theory to practice and to bring interesting research areas to the attention of conference participants. For the year 2001, the following three directions were chosen for presentation by the SOFSEM Steering Committee: – Trends in Informatics – Enabling Technologies for Global Computing – Practical Systems Engineering and Applications The above directions were covered through 12 invited talks presented by pro- nent researchers. There were 18 contributed talks, selected by the international Program Committee from among 46 submitted papers. The conference was also accompanied by workshops on Electronic Commerce Systems (coordinated by H. D. Zimmermann) and Soft Computing (coordinated by P. H´ajek).

关键词： Software Engineering/Programming and Operating Systems Theory of Computation computer Communication Networks Artificial Intelligence mathematics of Computing Information Systems and Communication Service

来源：评论

学校读者我要写书评

暂无评论

Simple and Efficient Graph Compression Schemes for Dense and Complement Graphs

引用

Journal of Combinatorial Optimization 1998年第4期2卷 351-359页

作者： Kao, Ming-Yang Occhiogrosso, Neill Teng, Shang-Hua Department of Computer Science Yale University New Haven CT 06520 United States Department of Computer Science Duke University Durham NC 27708 United States Department of Computer Science Univ. Illinois at Urbana-Champaign Urbana IL 61801 United States Department of Computer Science University of Minnesota Minneapolis MN 55455 United States Department of Mathematics Massachusetts Inst. of Technology Cambridge MA 02139 United States

We present two graph compression schemes for solving problems on dense graphs and complement graphs. They compress a graph or its complement graph into two kinds of succinct representations based on adjacency intervals and adjacency integers, respectively. These two schemes complement each other for different ranges of density. Using these schemes, we develop optimal or near optimal algorithms for fundamental graph problems. In contrast to previous graph compression schemes, ours are simple and efficient for practical applications.

关键词：

来源：评论

学校读者我要写书评

暂无评论

(Un)Expected Behavior of Digital Search Tree Profile 09

(Un)Expected Behavior of Digital Search Tree Profile

引用

Annual ACM-Society for Industrial and applied Mathmatics Symposium on Discrete Algorithms

作者： Michael Drmota Wojciech Szpankowski Inst. Discrete Mathematics and Geometry Department of Computer Science Purdue University

ISBN: (纸本)9780898716740

A digital search tree (DST) - one of the most fundamental data structures on words - is a digital tree in which keys (strings, words) are stored directly in (internal) nodes. Such trees find myriad of applications from the popular Lempel- Ziv'78 data compression scheme to distributed hash tables. The profile of a DST measures the number of nodes at the same distance from the root; it is a function of the number of stored strings and the distance from the root. Most parameters of DST (e.g., height, fill-up) can be expressed in terms of the profile. However, from the inception of DST, the analysis of the profile has been elusive and it has become a prominent open problem in the area of analysis of algorithms. We make here the first, but decisive, step towards solving this problem. We present a precise analysis of the average profile when stored strings are generated by a biased memoryless source. The main technical difficulty of analyzing the profile lies in solving a sophisticated recurrence equation. We present such a solution for the Poissonized version of the problem (i.e., when the number of stored strings is generated by a Poisson distribution) in the Mellin transform domain. To accomplish it, we introduce a novel functional operator that allows us to express the solution in an explicit form, and then using analytic algorithmics tools to extract the asymptotic behavior of the profile. This analysis is surprisingly demanding but once it is carried out it reveals unusually intriguing and interesting behavior. The average profile undergoes several phase transitions when moving from the root to the longest path. At first, it resembles a full tree until it abruptly starts growing polynomially and it oscillates in this range. Our results are derived by methods of analytic algorithmics such as generating functions, Mellin transform, Poissonization and de-Poissonization, the saddle-point method, singularity analysis and uniform asymptotic analysis.

关键词： Digital search trees Trees profile Analytic combinatorics Analysis of algorithms Generating functions Mellin transform

来源：评论

学校读者我要写书评

暂无评论

A Master Theorem for Discrete Divide and Conquer Recurrences 11

A Master Theorem for Discrete Divide and Conquer Recurrences

引用

Annual ACM-Society for Industrial and applied Mathmatics Symposium on Discrete Algorithms

作者： Michael Drmota Wojciech Szpankowski Inst. Discrete Mathematics and Geometry Department of Computer Science Purdue University

ISBN: (纸本)9780898719932

Divide-and-conquer recurrences are one of the most studied equations in computer science. Yet, discrete versions of these recurrences, namely T(n)=a_n+∑(b_jT([p_jn+δ_j])(1≤jcomputer science recurrence. We illustrate applicability of our results on several examples including a popular and fast arithmetic coding algorithm due to Boncelet for which we estimate its average redundancy. To the best of our knowledge, discrete divide and conquer recurrences were not studied in this generality and such detail; in particular, this allows us to com- pare the redundancy of Boncelet's algorithm to the (asymp-totically) optimal Tunstall scheme.

关键词： Divide-and-conquer recurrence Boncelet's data compression algorithm Dirichlet series Mellin-Perron formula Tauberian theorem Tauberian theorem Relapse Floor Function Divide and conquer computer science Theorem redundancy discrete element

来源：评论

学校读者我要写书评

暂无评论

Colour based road sign detection and extraction from still images

Colour based road sign detection and extraction from still i...

引用

IEEE International Conference on Cybernetic Intelligent Systems

作者： Malik, Rabia Nazir, Sana Khurshid, Javaid Department of Computer Science Military College of Signal NUST Rawalpindi Pakistan Elixir Technologies Corporation Software Technology Park F-5/1 Islamabad Pakistan Dept. of Computer and Info Sciences Pakistan Inst. of Engg. and Applied Science Islamabad Pakistan

ISBN: (纸本)9781424490233

This paper presents a system for the detection of road signs from a road scene image and extracts the pictogram inside the sign. The detection is carried out on basis of colour information of road signs inst.ad of edge information from grayscale images. The presented colour based detection process is invariant to the common problems in road signs i.e. shadow, highlight and general affine transformation and rotation. However owing to the problems addressed in the process, it can be stated that the colour information alone is inadequate for segmentation process. Using some edge based segmentation methodology alongside would increase the detection process' performance. The idea presented in the paper is through detection of road signs with red boundaries and white interior containing black pictograms. More colours with thresholds and shapes with their Hue moments can be added without any change in the general detection methodology;colour segmentation and shape analysis. HSV colour space is adopted for colour segmentation which makes detection invariable to high-lights and illumination changes. While affine moments are used for shape analysis, which makes it invariant to rotation and scale. © 2010 IEEE.

关键词： Fuzzy logic

来源：评论

学校读者我要写书评

暂无评论

Deep ViT Features as Dense Visual Descriptors

arXiv

引用

arXiv 2021年

作者： Amir, Shir Gandelsman, Yossi Bagon, Shai Dekel, Tali Dept. of Computer Science and Applied Math The Weizmann Inst. of Science United States

We study the use of deep features extracted from a pre-trained Vision Transformer (ViT) as dense visual descriptors. We observe and empirically demonstrate that such features, when extracted from a self-supervised ViT model (DINO-ViT), exhibit several striking properties, including: (i) the features encode powerful, well-localized semantic information, at high spatial granularity, such as object parts;(ii) the encoded semantic information is shared across related, yet different object categories, and (iii) positional bias changes gradually throughout the layers. These properties allow us to design simple methods for a variety of applications, including co-segmentation, part co-segmentation and semantic correspondences. To distill the power of ViT features from convoluted design choices, we restrict ourselves to lightweight zero-shot methodologies (e.g., binning and clustering) applied directly to the features. Since our methods require no additional training nor data, they are readily applicable across a variety of domains. We show by extensive qualitative and quantitative evaluation that our simple methodologies achieve competitive results with recent state-of-the-art supervised methods, and outperform previous unsupervised methods by a large margin. © 2021, CC BY.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

Knowledge translation: Computing the query potential of bio-ontologies

Knowledge translation: Computing the query potential of bio-...

引用

Workshop on Semantic Web Applications and Tools for Life sciences, SWAT4LS 2008

作者： Ang, Wee Tiong Kanagasabai, Rajaraman Baker, Christopher J.O. Data Mining Department Inst. for Infocomm Research 1 Fusionopolis Way Singapore 138632 Singapore Department of Computer Science and Applied Statistics University of New Brunswick Saint John NB Canada

Online ontologies have become a topic of discussion in a number of meta-data and content management communities including biology. For a number of technical and social reasons, domain experts are unable to get close enough to understand the conceptualization of an ontology they may wish to reuse or access as a query model. Consequentially they face initial conceptual challenges in how they can contribute to the ontology development process and what actual benefit they can derive from their contributions in the short and medium term. To ameliorate this need we report on the KnowleFinder system that summarizes queries that can be built from ontology as a query model and translates these into natural language statements for interpretation by the domain expert. Each natural language statement can then be submitted as an A-box reasoning query and its subsequent answer is retrieved. We illustrate the system with a subset of bio-ontologies.

关键词： Query processing

来源：评论

学校读者我要写书评

暂无评论

From reference frames to reference planes: Multi-view parallax geometry and applications 5th

From reference frames to reference planes: Multi-view parall...

引用

5th European Conference on computer Vision, ECCV 1998

作者： Irani, M. Anandan, P. Weinshall, D. Dept. of Applied Math and CS The Weizmann Inst. of Science Rehovot Israel Microsoft Research One Microsoft Way RedmondWA98052 United States Institute of Computer Science Hebrew University Jerusalem91904 Israel

ISBN: (纸本)3540646132

This paper presents a new framework for analyzing the geometry of multiple 3D scene points from multiple uncalibrated images, based on decomposing the projection of these points on the images into two stages: (i) the projection of the scene points onto a (real or virtual) physical reference planar surface in the scene;this creates a virtual "image" on the reference plane, and (ii) the re-projection of the virtual image onto the actual image plane of the camera. The positions of the virtual image points are directly related to the 3D locations of the scene points and the camera centers relative to the reference plane alone. All dependency on the internal camera calibration parameters and the orientation of the camera are folded into homographies relating each image plane to the reference plane. Bi-linear and tri-linear constraints involving multiple points and views are given a concrete physical interpretation in terms of geometric relations on the physical reference plane. In particular, the possible dualities in the relations between scene points and camera centers are shown to have simple and symmetric mathematical forms. In contrast to the plane+parallax (p+p) representation, which also uses a reference plane, the approach described here removes the dependency on a reference image plane and extends the analysis to multiple views. This leads to simpler geometric relations and complete symmetry in multi-point multiview duality. The simple and intuitive expressions derived in the reference-plane based formulation lead to useful applications in 3D scene analysis. In particular, simpler tri-focal constraints are derived that lead to simple methods for New View Synthesis. Moreover, the separation and compact packing of the unknown camera calibration and orientation into the 2D projection transformation (a homography) allows also partial reconstruction using partial calibration information. © Springer-Verlag Berlin Heidelberg 1998.

关键词： Geometry

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：