检索结果-内蒙古大学图书馆

您好，读者！请登录

内蒙古大学图书馆

首页
概况
党建
资源
服务
科研支持
- 论文收录引用证明
- 科技查新
知识产权
档案馆
帮助

咨询与建议

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

您的常用邮箱：*

您的手机号码：*

问题描述：

当前已输入0个字，您还可以输入200个字

全部搜索
期刊论文
图书
学位论文
标准
纸本馆藏
外文资源发现
数据库导航
超星发现

高级检索

时间限定

出版年份：

文献类型

图书期刊文献学位论文多媒体

馆藏选择

电子馆藏纸本馆藏

核心期刊

全部期刊 SCI 收录期刊 SSCI 收录期刊 EI 收录期刊 CSCD 收录期刊 CSSCI 收录期刊

语言

中文英文

文献类型

期刊文献图书学位论文标准纸本馆藏

帮助

文字说明：

T=题名（书名、题名），A=作者（责任者），K=主题词，P=出版物名称，PU=出版社名称，O=机构（作者单位、学位授予单位、专利申请人），L=中图分类号，C=学科分类号，U=全部字段，Y=年（出版发行年、学位年度、标准发布年）

检索规则说明：

AND代表“并且”；OR代表“或者”；NOT代表“不包含”；(注意必须大写,运算符两边需空一格)

检索范例：

范例一：(K=图书馆学 OR K=情报学) AND A=范并思 AND Y=1982-2016
范例二：P=计算机应用与软件 AND (U=C++ OR U=Basic) NOT K=Visual AND Y=2011-2016

分类表

所选分类

>> <<

限定检索结果

文献类型

136 篇 会议

馆藏范围

136 篇 电子文献
0 种 纸本馆藏

日期分布

学科分类号

75 篇 理学
- 75 篇 系统科学
32 篇 工学
- 15 篇 计算机科学与技术...
- 14 篇 软件工程
- 7 篇 机械工程
- 7 篇 信息与通信工程
- 5 篇 控制科学与工程
- 2 篇 仪器科学与技术
- 1 篇 网络空间安全
28 篇 文学
- 25 篇 中国语言文学
- 3 篇 外国语言文学
16 篇 法学
- 10 篇 社会学
- 6 篇 民族学
2 篇 经济学
- 2 篇 应用经济学
2 篇 管理学
- 2 篇 管理科学与工程(可...
2 篇 艺术学
- 2 篇 设计学（可授艺术学...

主题

15 篇 语音识别
4 篇 语种识别
4 篇 计算机辅助语言学...
3 篇 中国科学院自动化...
3 篇 frame
3 篇 ncmmsc2009
3 篇 微软亚洲研究院
3 篇 语音合成
3 篇 语言模型
3 篇 电子工程系
2 篇 说话人确认
2 篇 检索机
2 篇 鲁棒性
2 篇 ncmmsc
2 篇 置信度
2 篇 哼唱识别
2 篇 语音学
2 篇 元音
2 篇 情感识别
2 篇 模糊匹配

机构

19 篇 清华大学
5 篇 西北民族大学
5 篇 中国传媒大学
4 篇 北京交通大学
4 篇 中国社会科学院语...
3 篇 ibm中国研究院
3 篇 哈尔滨工业大学
3 篇 北京理工大学
3 篇 ibm中国研究中心
2 篇 清华信息科学技术...
2 篇 太原理工大学
2 篇 national institu...
2 篇 北京邮电大学
2 篇 清华-讯飞语音技术...
2 篇 北京语言大学
2 篇 中国科学院自动化...
2 篇 graduate school ...
2 篇 北京信息科技大学
2 篇 中国科学技术大学
2 篇 清华大学信息技术...

作者

7 篇 郑方
6 篇 秦勇
5 篇 蔡莲红
5 篇 于洪志
4 篇 邬晓钧
4 篇 李爱军
4 篇 徐明星
3 篇 谢湘
3 篇 苗振江
3 篇 刘轶
3 篇 夏云庆
3 篇 徐波
3 篇 吴及
3 篇 孟子厚
3 篇 双志伟
3 篇 戴礼荣
2 篇 王士进
2 篇 张雪英
2 篇 吐尔洪江·阿布都克...
2 篇 匡镜明

语言

90 篇 中文
46 篇 英文

检索条件"任意字段=第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会"

共 136 条记录，以下是121-130 订阅

全选清除本页清除全部题录导出标记到"检索档案"

详细简洁

排序：

So You Want to Design an Intelligent Communication Machine

So You Want to Design an Intelligent Communication Machine

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： B.H.Juang Georgia Institute of Technology

Advances in modern technologies have intensified the interest in injecting human intelligent capabilities to mechanical automation and its realization. One of the most obvious intelligent capabilities or its demonstration is for a machine to communicate in a humanly fashion. This appears to be the theme of the conference. In this talk, we shall first look at an interesting historical aspect of the so-called intelligence, namely the idea of developing a mathematical framework for measuring and processing "intelligence." Such a discussion, although non-conclusive, will help to lay a possible foundation for what might be considered an analytical approach to machine intelligence. The second part of the talk will focus on the process of speech communication itself. We use a layered view of the process, with corresponding technologies under development as opposed to a traditional taxonomy based on linguistic theory, to enunciate the way human-like intelligence can be introduced and incorporated into a conversing machine, beyond the simple act of recognizing spoken words. Such a technology-based view of the speech communication process will be highlighted in a video demonstration.

关键词： So You Want to Design an Intelligent Communication Machine

来源：评论

学校读者我要写书评

暂无评论

Speech Source in Speech Synthesis

Speech Source in Speech Synthesis

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Kong Jiangping Dept. of Chinese and Research Center of Chinese Linguistics University,China HKPEAC, Faculty of Education, the University of Hong Kong,Hong Kong

This paper is concerned with the study on speech source and speech synthesis through the methods of acoustics, EGG and dynamic glottis, and the method of formant speech synthesis. From the viewpoint of phonetics and speech science, speech production includes three aspects which are speech source, speech resonance and lip radiation. But now the methods in the speech engineering usually do not following the principle of production, because we are lack of knowledge of the speech production. In this research, we studied the speech source through acoustical signal, EGG signal and high-speed image signal and then main parameters which are of the main properties of speech source are extracted for synthesis. Depending on the research, a formant synthesis system has been established for improving the properties of speech source and speech source models. According to our study, we have set up a dynamic glottis model in the synthesis system, which can be used for the purpose of the speech research and speech singineering.

关键词： frame image Speech Source in Speech Synthesis

来源：评论

学校读者我要写书评

暂无评论

Speech and emotion transformation based on stable frame features of phonemes

Speech and emotion transformation based on stable frame feat...

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Xiang Wang, Jianping Zhang, Yonghong Yan Thinkit Speech Lab, Institute of Acoustics, Chinese Academy of Sciences, Beijing 100080, China

We present our current research on Speech and emotional Transformation based on stable frame features of phonemes. Speech transformation is an effective way to transform an original speaker's voice to sound as if it is uttered by the target speaker and emotional voice conversion is needed in commercial TTS systems to provide a user-friendly interface for customers. Our speech transformation algorithms is introduced which involves threes phases. First, the acoustic feature distributions for source speaker and target speaker are modeled using Gaussian Mixture Model. Second, the Expectation-Maximum algorithm is exploited to optimize the model parameters. Finally, an MMSE criterion is used to get the transformation formula for converting the acoustic features from source speaker to target speaker. The database consisting of five kinds of speech emotions (i.e. happiness, sadness,surprise,anger,neutral). Perceptual experimental results are given and some analysis is used in our emotional conversion experiments.

关键词： TTS Speech and emotion transformation based on stable frame features of phonemes frame

来源：评论

学校读者我要写书评

暂无评论

BOOST GRAPHEME-TO-PHONEME CONVERSION FOR MANDARIN TTS

BOOST GRAPHEME-TO-PHONEME CONVERSION FOR MANDARIN TTS

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Lifu Yi1, Jian Li1, Jie Hao1, Ziyu Xiong2 1Toshiba (China) Research and Development Center, Beijing, 100086 2Institute of Linguistics, Chinese Academy of Social Sciences, Beijing, 100732

In this paper, several methods are introduced to boost grapheme-to-phoneme (G2P) conversion models for Chinese text-to-speech (TTS) systems. We try to ease the critical problem of data sparsity with combined approaches. Firstly, we introduce a text-design method to cover more contexts of G2P text corpus as many as possible. Secondly, we use various data-driven methods for modeling, and then we make comparisons to select the best method for each polyphonic word. Finally, we build independent models for some neutral tone words besides the normal G2P models, which help to achieve more compact and flexible G2P models. The experiments show that these methods yield more than 50% relative error reduction both for normal polyphonic words and Chinese neutral tones.

关键词： Grapheme-to-phoneme conversion text design Chinese neutral tone speech synthesis

来源：评论

学校读者我要写书评

暂无评论

Spontaneous Speech Database Design for Suprasegmental Characteristics in Asian L2 English

Spontaneous Speech Database Design for Suprasegmental Charac...

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Tanya Visceglia Chiu-yu Tseng Department of Applied English Ming Chuan University

This research is part of the ongoing multinational collaboration "Asian English Speech cOrpus Project" (AESOP), whose aim is to build up a speech corpus representing the varieties of English spoken in Asia. The present paper describes tasks designed to elicit production of a comprehensive range of English segmental and suprasegmental characteristics in the form of spontaneous speech. Segmental and suprasegmental properties of spontaneous speech have been shown to differ significantly from those of read speech;nevertheless, much of the data used to develop man-machine speech communication systems has, for the most part, been based on read speech. Experiments designed to elicit a full range of L2 English segmental and prosodic features in spontaneous speech could efficiently collect an inventory of these features in a database, which could be used for further phonetic studies as well as modeling and ICT tool development tailored to the Asian L2 English-speaking population.

关键词： L2 English, Asia, speech corpus, spontaneous speech, computer-prompted dialogue, picture description, discourse prosody, suprasegmental features

来源：评论

学校读者我要写书评

暂无评论

Mandarin Infants’ Familiar Word Comprehension

Mandarin Infants’ Familiar Word Comprehension

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Twila Tardif Rachel Pulverman Institute of Psychology Chinese Academy of Sciences、University of Michigan

Purpose: The present study examines the comprehension of familiar words that children acquired early in the terms of object vs. action among Mandarin-speaking infants. Method: An IPLP(intermodal preferential-looking Paradigm)was adopted that infants, ranging in age from 11 months,22 days to 24 months, 29 days, grouped as 14-, 18- and 24-month-olds, view split-screen 'preference' scenes in which the agent performs a familiar action on a novel object in one scene (e.g., Po kissing a ball) and, simultaneously, in the other scene, the agent performs a novel action on the familiar object. Infants' looking time of target word and non-target word was recorded. Result: 24-month-olds showed true understanding of both the familiar nouns and verbs, while 14- and 18- month-olds watch the target side and non-target side randomly. All the three groups (14-, 18-, and 24-month-olds) of infants in the study showed better understanding of nouns than verbs.

关键词： infant word learning culture

来源：评论

学校读者我要写书评

暂无评论

Pronunciation Modeling for Nonstandard Speech Using Phonetic Feature Distance and Optimal Gaussian Mixture Sharing

Pronunciation Modeling for Nonstandard Speech Using Phonetic...

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Yi Liu 1, Yongsheng Yang 2, Shilei Huang 2,3, Wenxiao Cao 1 1. Tsinghua University 2. HKUST-IER Intelligent Media and Speech Lab 3. Hong Kong University of Science and Technology

Pronunciations in nonstandard speech that including causal speech, accented speech, spontaneous and conversational speech tend to be much more varied than in carefully read speech. ASR systems can reach an accuracy of above 90% when evaluated on read speech, but the accuracy of nonstandard speech is much lower. This high error rate is due in part to the poor modeling of the diversity of pronunciation variations within nonstandard speech. In this paper, we propose modeling the variations at iterative acoustic model training as well as acoustic model generation levels. We describe our method of incorporating phonetic feature distance into the calculation of phone variation probabilities. We also present an efficient criterion for choosing the optimal Gaussian mixture components from surface states, and share the selected Gaussian mixture components with canonical states according to variation probabilities. Experiments have shown that phonetic feature distance based state level Gaussian mixture sharing gives a significant syllable error rate reduction of 3.05% absolutely with respect to the baseline.

关键词： Nonstandard Speech Phonetic Feature Distance Gaussian Mixture Sharing

来源：评论

学校读者我要写书评

暂无评论

Visualization Method of Similarities among Multiple Speech Corpora

Visualization Method of Similarities among Multiple Speech C...

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Kimiko YAMAKAWA Tomoko MATSUI Shuichi ITAHASHI National Institute of Informatics (NII) National Institute of Advanced Industrial Science and Technology (AIST)

The purpose of this study is to visualize the similarities among multiple speech corpora. Speech data are indispensable for promoting speech research. A wide variety of speech corpora has recently been developed in many countries. Corpus diversification has given users many choices for corpus selection. In order for users to easily utilize these various corpora, we proposed a visualization method based on the corpus attribute using MDS. First, we listed eight attributes of the speech corpora. Then, we selected a few items for each attribute resulting in 58 items in all. Each item takes on a '1' or '0' depending on whether the corpus has the attribute or not. The set of corpus features is represented as a 58-dimensional vector. Then, the vectors are converted into a similarity matrix and analyzed using a multidimensional scaling method (MDS). We analyzed the speech corpora distributed by the peech Resources Consortium (NII-SRC). The results showed that it is possible to visualize the similarities among multiple speech corpora using the proposed method. This will facilitate searching efficiently the specific corpus that fits a user's needs.

关键词： MDS Visualization Method of Similarities among Multiple Speech Corpora

来源：评论

学校读者我要写书评

暂无评论

Rudimentary Form Class Categorization of Nouns and Verbs in Chinese-Learning Infants

Rudimentary Form Class Categorization of Nouns and Verbs in ...

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： ZHANG Zhao*, SHI Rushen**, LI Aijun* & ZHENG Leilei*** *Institute of Linguistics, Chinese Academy of Social Sciences **Université du Québec à Montréal ***Shandong Normal University

We hypothesized that infants can use distributional information to categorize words to basic noun- and verb-like classes. In a visual procedure, 11-16-month-old Chinese-learning infants were familiarized with 3 novel words in function word contexts appropriate for verbs (我也, 你别), and 3 other novel words in noun contexts (我的, 这个). The pairings of the novel words with the noun versus verb contexts were counterbalanced across infants. One word in one noun context and one word in one verb context were withheld during familiarization. Infants were then tested with these withheld phrases. In grammatical trials, the words occurred with the correct withheld nominal and verbal contexts. In ungrammatical trials, the same words crossed the contexts, violating the familiarized distributional patterns. Looking times were significantly different for the two test trial types. Conclusion: Chinese infants can use purely distributional information to categorize novel words to rudimentary noun and verb classes.

关键词： Rudimentary Form Class Categorization of Nouns and Verbs in Chinese-Learning Infants

来源：评论

学校读者我要写书评

暂无评论

Data alignment for text-independent voice conversion

Data alignment for text-independent voice conversion

引用

第十届全国人机语音通讯学术会议暨国际语音语言处理研讨会

作者： Jianhua Tao, Meng ZhangNational Laboratory of Pattern Recognition, Institute of Automation, Chinese Academyy of Sciences

Voice conversion has become more and more important in speech technology, however most of current works have to use parallel corpus including both source and target speaker as the training corpus, that limits the application of the technology. In the talk, we present a method of text-independent voice conversion which uses non-parallel corpus for the training. We use phonetic information for data alignment of non-parallel corpus. HMM is used for representing the phonetic structure of training speech. States belong to same phoneme are considered together as a phoneme cluster. By the adaptation of corresponding phoneme clusters from source and target speech, state mapped codebook is established using weighted linear transform. For each source vector, several nearest clusters are considered simultaneously while mapping in order to generate a continuous and stable transform. Then, the text-independent voice conversion system is finally created. Final subjective experiments based on ABX tests and MOS tests show that the proposed method gets the similar converting performance and better speech quality compared to the conventional voice conversion systems. The proposed technique is also applicable to both intra-lingual and cross-lingual voice conversion.

关键词： HMM Data alignment for text-independent voice conversion

来源：评论

学校读者我要写书评

暂无评论

没有更多数据了...

全选清除本页清除全部题录导出标记到“检索档案”

共14页 << < 5 6 7 8 9 10 11 12 13 14 > >>

检索报告对象比较合并检索0

隐藏清空

合并搜索

回到顶部

执行限定条件

内容：

评分：

请选择保存的检索档案：

请选择收藏分类：

订阅名称：

通借通还

温馨提示：

图书名称：

借书校区：

取书校区：

手机号码：

邮箱地址：

一卡通帐号：

电话和邮箱必须正确填写，我们会与您联系确认。

联系人：

所在院系：

联系邮箱：

联系电话：

内蒙古自治区呼和浩特市赛罕区大学西街235号邮编: 010021

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：