课题基金 / 基金详情

SGER: Multi-Stream Approach Using Syllable Length Temporal Evidence in Acoustic Modeling of Conversational Speech

SGER: Multi-Stream Approach Using Syllable Length Temporal Evidence in Acoustic Modeling of Conversational Speech
SGER:在会话语音声学建模中使用音节长度时间证据的多流方法
批准号:
9713583
负责人:
Hynek Hermansky
金额:
$4.97万
依托单位国家:
美国
项目类别:
Standard Grant
财政年份:
1997
资助国家:
美国
项目状态:
已结题
起止时间:
1997-07-01 至 1998-12-31

项目摘要

项目成果

Hynek Hermansky的其他基金

相似基金

相关文献

中文摘要
翻译
该项目研究多流方法自动语音识别(ASR)使用流定义在调制频谱域。多流方法是一种新的声学建模技术,可以提高ASR系统的鲁棒性。它从语音信号中提取不同的信息子流,然后在每个子流中进行独立的特征提取和概率估计。将各子流中得到的类条件概率向量进行合并,得到最终决策。利用子流中频谱的不同区域的多频带模型已被证明对频率选择性退化具有鲁棒性。该项目的目的是在不同的子流中利用不同范围的调制频谱。对调制频谱不同范围的相对重要性的研究表明,它们可能携带不同的信息。这使得调制频谱的各种元素成为多流方法的有趣候选者。这些特征的计算需要查看较长的信号片段(通常是音节长度的片段)。该项目试图将特征向量元素的相对独立性和特征向量之间的中期时间依赖性的新假设应用于ASR,以提高其鲁棒性。
英文摘要
The project studies the multi-stream approach to automatic speech recognition (ASR) using stream definitions in the modulation spectrum domain. The multi-stream approach is a new acoustic modeling technique to increase the robustness of ASR systems. It derives different information sub-streams from the speech signal followed by independent feature extraction and probability estimation in each sub-stream. The vectors of class-conditional probabilities obtained from each sub-stream are merged to obtain the final decision. The multi-band model which utilizes different regions of the frequency spectrum in the sub-streams has been shown to be robust to frequency-selective degradation. The aim of the project is to utilize different ranges of the modulation spectrum in different sub-streams. A study of the relative importance of the different ranges of the modulation spectrum indicates that they might carry different information. This makes the various elements of the modulation spectrum interesting candidates for the multi-stream approach. The computation of these features requires looking at longer segments (typically syllable length segments) of the signal. The proposed project is an attempt to apply the new assumptions of relative independence of elements of the feature vector and medium-term time dependencies between feature vectors to ASR with a view to improve its robustness.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
Deriving Hearing Knowledge from Speech Data
  • 批准号:
    1743616
  • 项目类别:
    Standard Grant
  • 资助金额:
    $7.99万
  • 财政年份:
    2017
  • 负责人:
    Hynek Hermansky
  • 依托单位:
国内基金
海外基金
基于Multi-Pass Cell的高功率皮秒激光脉冲非线性压缩关键技术研究
Multi-decadeurbansubsidencemonitoringwithmulti-temporaryPStechnique
  • 批准号:
    --
  • 项目类别:
    --
  • 资助金额:
    80万元
  • 批准年份:
    2022
  • 负责人:
    Timo Balz
  • 依托单位:
High-precision force-reflected bilateral teleoperation of multi-DOF hydraulic robotic manipulators
  • 批准号:
    52111530069
  • 项目类别:
    国际(地区)合作与交流项目
  • 资助金额:
    10万元
  • 批准年份:
    2021
  • 负责人:
    徐兵
  • 依托单位:
大地电磁强噪音压制的Multi-RRMC技术及其在青藏高原东南缘-印支块体地壳流追踪中的应用