专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

81. 发明申请

WO00045374A1 A METHOD AND PORTABLE APPARATUS FOR PERFORMING SPOKEN LANGUAGE TRANSLATION 审中-公开
标题翻译：一种用于执行语音翻译的方法和便携式设备
公开(公告)号：WO00045374A1
公开(公告)日：2000-08-03
申请号：PCT/US1999/027874
申请日：1999-11-24
IPC分类号： G06F17/27 , G06F17/28 , G10L15/06 , G10L15/20
CPC分类号： G06F17/2765 , G06F17/271 , G06F17/2755 , G06F17/2827 , G06F17/2872 , G10L15/07 , G10L15/20
摘要： In a portable unit, a method for performing spoken language translation. The method includes the steps of receiving (302) at least one speech input comprising at least one source language, and recognizing (306) at least one source expression of the at least one source language. The method also includes the steps of translating (308) the recognized at least one source expression from the at least one source language to at least one target language, synthesizing (312) at least one speech output from the translated at least one target language, and providing (314) the at least one speech output. A set of source language expressions and a set of target language expressions are remotely extensible in some embodiments, using various communication methods. One embodiment further comprises the step of minimizing (1702) misrecognitions of the at least one source expression, wherein the misrecognitions result from factors comprising noise and speaker variation.
摘要翻译：在便携式单元中，执行口语翻译的方法。该方法包括以下步骤：接收（302）包括至少一种源语言的至少一个语音输入，以及识别（306）所述至少一种源语言的至少一个源表达式。该方法还包括将识别的至少一个源表达从至少一个源语言翻译（308）到至少一个目标语言的步骤，从翻译的至少一个目标语言合成（312）至少一个语音输出，以及提供（314）所述至少一个语音输出。使用各种通信方法，在一些实施例中，源语言表达式和一组目标语言表达式可远程扩展。一个实施例还包括最小化（1702）至少一个源表达式的错误识别的步骤，其中由包括噪声和说话者变化的因素导致误识别。

82. 发明申请

WO00036592A1 IMPROVED NOISE SPECTRUM TRACKING FOR SPEECH ENHANCEMENT 审中-公开
标题翻译：改进的噪声跟踪用于语音增强
公开(公告)号：WO00036592A1
公开(公告)日：2000-06-22
申请号：PCT/US1999/029901
申请日：1999-12-16
IPC分类号： G10L15/20 , G10L21/02
CPC分类号： G10L21/0208 , G10L15/20 , G10L21/0216 , G10L2021/02168
摘要： A spectrum-based speech enhancement system estimates and tracks the noise spectrum of a mixed speech and noise signal. The system frames and windows a digitized signal and applies the frames to a fast Fourier transform processor to generate discrete Fourier transformed (DFT) signals representing the speech plus noise signal. The system calculates the power spectrum of each frame. The speech enhancement system employs a leaky integrator that is responsive to identified noise-only components of the signal. The leaky integrator has an adaptive time-constant which compensates for non-stationary environmental noise. In addition, the speech enhancement system identified noise-only intervals by using a technique that monitors the Teager energy of the signal. The transition between noise-only signals and speech plus noise signals is softened by being made non-binary. Once the noise spectrum has been estimated, it is used to generate gain factors that multiply the DFT signals to produce noise-reduced DFT signals. The gain factors are generated based on an audible noise threshold. The method generates audible a priori and a posteriori signal to noise ratio signals and then calculates audible gain signals from these values.
摘要翻译：基于频谱的语音增强系统估计和跟踪混合语音和噪声信号的噪声谱。系统帧和窗口是数字化信号，并将帧应用于快速傅里叶变换处理器，以产生表示语音加噪声信号的离散付里叶变换（DFT）信号。系统计算每帧的功率谱。语音增强系统采用泄漏积分器，其响应于所识别的仅噪声信号分量。泄漏积分器具有补偿非平稳环境噪声的自适应时间常数。此外，语音增强系统通过使用监视信号的Teager能量的技术来识别仅噪声间隔。仅噪声信号和语音加噪声信号之间的转换通过非二进制软化。一旦已经估计出噪声频谱，它被用于产生乘以DFT信号以产生噪声降低的DFT信号的增益因子。基于可听噪声阈值产生增益因子。该方法产生可听的先验和后验信噪比信号，然后从这些值计算可听增益信号。

83. 发明申请

WO00031728A1 SPEECH RECOGNITION METHOD IN A NOISY ACOUSTIC SIGNAL AND IMPLEMENTING SYSTEM 审中-公开
标题翻译：语音信号和实现系统中的语音识别方法
公开(公告)号：WO00031728A1
公开(公告)日：2000-06-02
申请号：PCT/FR1999/002852
申请日：1999-11-19
IPC分类号： G10L15/20
CPC分类号： G10L15/20
摘要： The invention concerns a method and a system for speech recognition in a noisy acoustic signal. In a preferred embodiment, the system (2) comprises modules (30) for detecting speech and for producing a noise model (31), a module (40) quantifying the noise energy level and comparing with predetermined energy level ranges, a parametering chain (5) comprising an optional module for noise correction (51), with Wiener filter, a module (52) for calculating spectral energy in Bark windows, a module (50, 530) for applying a configuration of shifted values (531), by adding said values to Bark coefficients, based on the quantification (40) to modify the parameterization, a module (54) for calculating parameter vectors, and a unit (6) for pattern recognition, performing speech recognition by comparison with pre-recorded parameter vectors during a learning phase.
摘要翻译：本发明涉及一种用于噪声声信号中的语音识别的方法和系统。在优选实施例中，系统（2）包括用于检测语音和用于产生噪声模型（31）的模块（30），对噪声能级进行量化并与预定能级范围比较的模块（40），参数链（ 5）包括用于噪声校正的可选模块（51），具有维纳滤波器，用于计算Bark窗口中的频谱能量的模块（52），用于通过添加移位值（531）来构造的模块（50,530）基于用于修改参数化的量化（40），用于计算参数矢量的模块（54）和用于模式识别的单元（6），将所述值表示为Bark系数，通过与预先记录的参数向量进行比较来执行语音识别学习阶段

84. 发明申请

WO99040570A1 METHOD AND DEVICE FOR ENHANCING RECOGNITION PROBABILITY IN VOICE RECOGNITION SYSTEMS 审中-公开
标题翻译：方法和设备用于提高语音识别系统的可能性检测
公开(公告)号：WO99040570A1
公开(公告)日：1999-08-12
申请号：PCT/DE1999/000137
申请日：1999-01-20
IPC分类号： G10L15/06 , G10L15/20 , G10L5/06
CPC分类号： G10L15/063 , G10L2015/0635
摘要： The invention relates to a method for enhancing recognition probability in voice recognition systems. According to the inventive method, selective post-training of the already stored homonymic term is carried out after inputting a term to be recognized. This makes it possible to improve the speaker-dependent recognition rate even in environments with prevailing acoustic interference.
摘要翻译：本发明涉及提高语音识别系统的识别概率的方法。该方法是基于术语的该输入后要识别被执行，可选地，同名已经存储项的再训练。这确保了说话者相关的识别率，即使在声学噪声环境实现改善。

85. 发明申请

WO99014739A1 METHOD FOR SUPPRESSING NOISE IN A DIGITAL SPEECH SIGNAL 审中-公开
标题翻译：用于在数字语音信号中抑制噪声的方法
公开(公告)号：WO99014739A1
公开(公告)日：1999-03-25
申请号：PCT/FR1998/001981
申请日：1998-09-16
IPC分类号： G10L15/20 , G10L15/02 , G10L21/02 , G10L21/0208 , G10L21/0232 , G10L21/0264 , G10L25/90 , H04B3/20 , G10L3/02
CPC分类号： G10L21/0208 , G10L21/0232 , G10L21/0264 , G10L21/0364 , G10L25/90
摘要： The invention concerns a method for suppressing noise in a digital speech signal processed by successive frames which consists in: computing the signal spectral components (Sn,f, Sn,i) on each frame; computing the maximised estimations (B'?n,i?) of spectral components of the noise included in the speech signal; carrying out a harmonic analysis of the signal to estimate a pitch; carrying out a spectral subtraction comprising at least a step consisting in subtracting respectively, from each spectral component of the speech signal on the frame (Sn,f), a quantity depending on parameters including the maximised estimation of the noise corresponding spectral component and the estimated pitch; and applying to the subtraction result a transform towards the time domain to construct an enhanced speech signal (s ).
摘要翻译：本发明涉及一种用于抑制由连续帧处理的数字语音信号中的噪声的方法，其中包括：计算每帧上的信号频谱分量（Sn，f，Sn，i）计算包括在语音信号中的噪声的频谱分量的最大化估计（B'？n，i？）; 对信号进行谐波分析以估计音高; 执行频谱减法，其包括至少包括从帧（Sn，f）上的语音信号的每个频谱分量中减去取决于包括噪声对应频谱分量的最大化估计的参数的量的步骤，以及估计沥青; 并且向减法结果应用向时域的变换以构建增强的语音信号（s 3）。

86. 发明申请

WO99003091A1 METHODS AND APPARATUS FOR MEASURING SIGNAL LEVEL AND DELAY AT MULTIPLE SENSORS 审中-公开
标题翻译：用于在多个传感器上测量信号电平和延迟的方法和装置
公开(公告)号：WO99003091A1
公开(公告)日：1999-01-21
申请号：PCT/SE1998/001319
申请日：1998-07-03
IPC分类号： G10L15/28 , G10K11/34 , G10L15/20 , G10L21/02 , H04B1/10 , H04M9/08 , H04R3/00 , G10K11/178
CPC分类号： G10K11/345 , H04M9/08
摘要： Methods and apparatus for automatic gain and time delay computations between multiple sensors. In exemplary embodiments, one sensor is treated as a reference sensor with respect to a measured signal quantity which is assumed to be a narrowband process. The relative gain and relative time delay in the additional sensors with respect to the same measured quantity are then automatically calculated based on an adaptive filtering algorithm. Advantageaously, the disclosed embodiments are implemented using standard digital signal processing components.
摘要翻译：多传感器之间自动增益和时间延迟计算的方法和装置。在示例性实施例中，相对于被假设为窄带处理的测量信号量，将一个传感器视为参考传感器。然后，基于自适应滤波算法自动计算附加传感器中相对于相同测量量的相对增益和相对时间延迟。有利地，所公开的实施例使用标准数字信号处理组件来实现。

87. 发明申请

WO98039946A1 DEVICE AND METHOD FOR PROCESSING SPEECH 审中-公开
标题翻译：用于处理语音的设备和方法
公开(公告)号：WO98039946A1
公开(公告)日：1998-09-11
申请号：PCT/JP1998/000915
申请日：1998-03-05
IPC分类号： G10L15/20 , G10L21/02 , H03H17/02 , H04B3/23 , H04M9/08 , H04R3/00 , H04R3/02
CPC分类号： H04R3/02 , G10L15/20 , G10L2021/02082 , H04M9/082
摘要： A speech processor which continuously uses the impulse response used by the preceding frame when a speech signal is included in the microphone input signal or the currently updated impulse response when no speech signal is included in the microphone input signal as the impulse response used for generating a pseudo-acoustic echo signal at the time of performing echo cancellation by using the pseudo-acoustic echo signal.
摘要翻译：当语音信号包括在麦克风输入信号中时，连续使用由前一帧使用的脉冲响应的语音处理器，或者当麦克风输入信号中没有语音信号被包括在麦克风输入信号中时当前更新的脉冲响应，作为用于产生通过使用伪声回波信号进行回波消除时的伪声回波信号。

88. 发明申请

WO9725790A3 NOISE CANCELLATION AND NOISE REDUCTION APPARATUS 审中-公开
标题翻译：噪声消除和噪声减少装置
公开(公告)号：WO9725790A3
公开(公告)日：1998-01-15
申请号：PCT/IB9601506
申请日：1996-06-04
申请人： ANDREA ELECTRONICS CORP
发明人： ANDREA DOUGLAS , TOPF MARTIN
IPC分类号： G10L15/20 , G10K11/16 , G10K11/178 , G10L21/02 , H04B1/12 , H04M1/03 , H04M1/05 , H04M1/19 , H04M1/60 , H04M1/74 , H04R1/04 , H04R1/10 , H04R3/00 , H04R3/02 , H04R29/00 , H04B15/00
CPC分类号： H04R1/1083 , G10K11/178 , H04B1/12 , H04M1/05 , H04M1/19 , H04M1/6008 , H04M9/08 , H04R1/10 , H04R1/1008 , H04R3/00 , H04R3/005 , H04R5/0335 , H04R29/006 , H04R2201/107 , H04R2201/403
摘要： This invention relates to a method and an apparatus for reducing ambient noise for use with a headset or a boom headset attached to a boom microphone device or the like. The apparatus can include a sensor microphone to detect a background noise signal (MIC 14), a desired input audio transmission (MIC 12), and signal processing means (16) for canceling the noise signals to create an inverted antinoise signal within an acoustical waveguide located adjacent to the earphone of headset. The method for reducing noise according to this invention is provided by an open loop circuit allowing the input audio signal from an operator or caller ('FROM TELEPHONE LINES') to be transmitted to the user's ear without the disturbance of unwanted ambient noise.

89. 发明申请

WO2023074655A1 情報処理装置、情報処理方法、プログラムおよび記録媒体审中-公开
公开(公告)号：WO2023074655A1
公开(公告)日：2023-05-04
申请号：PCT/JP2022/039618
申请日：2022-10-25
申请人：パイオニア株式会社
发明人：河内　洋人 , 中川　壮志
IPC分类号： G10L21/0208 , G10L15/20
摘要：ノイズを除去するためのフィルタ特性の算出処理の実行頻度を下げる情報処理装置を提供する。　収音部１０において収音した音声に基づいて、音声から雑音を除去するためのフィルタ特性を算出するフィルタ特性算出部３０と、フィルタ特性算出部３０によって算出されたフィルタ特性に基づいて、収音した音声から雑音を除去するフィルタ部２０と、センサ部４０のセンサ情報に基づいて音声の収音環境を検出する収音環境検出部５０と、収音環境検出部５０において検出された収音環境の変化を検出したとき、フィルタ特性算出部３０においてフィルタ特性を算出させ、該フィルタ特性をフィルタ部２０に設定するフィルタ制御部６０と、を備えている。

90. 发明申请

WO2022227365A1 基于词库学习的饱和潜水氦语音解读方法审中-公开
公开(公告)号：WO2022227365A1
公开(公告)日：2022-11-03
申请号：PCT/CN2021/116054
申请日：2021-09-01
申请人：南通大学
发明人：张士兵 , 吴建绒 , 郭莉莉 , 李明 , 包志华
IPC分类号： G10L15/20 , G10L15/22
摘要：一种基于词库学习的饱和潜水氦语音解读方法，该方法应用于包括潜水员、校正网络和解读网络的系统中，该方法包括：建立饱和潜水作业常用工作语言词库，由潜水员分别在不同环境下朗读，生成校正网络的监督信号和矢量信号；校正网络对不同潜水员在不同潜水深度的氦语音进行学习得到网络校正参数，并对潜水员的氦语音进行校正得到校正语音；解读网络对校正语音进行学习并完成氦语音的解读。该方法通过设置潜水员在不同环境下个体语音信号特征和工作语言词库特征，将机器学习引入到氦语音解读中，解决了饱和潜水氦语音的解读问题，减少了网络对学习样本数的要求，提高了氦语音解读器在不同潜水深度的自适应性，尤其适合于不同潜水深度环境下的氦语音的完备解读。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式