公开(公告)号：WO1983003701A1

公开(公告)日：1983-10-27

申请号：PCT/US1983000461

申请日：1983-04-04

Applicant: EAR THREE SYSTEMS MANUFACTURING COMPANY

Inventor： EAR THREE SYSTEMS MANUFACTURING COMPANY ,  DuBRUCQ, Denyse, C.

IPC: G10L01/00

CPC classification number: G09B21/009

Abstract: A simulated speech method and apparatus afford a dual matrix presentation (32 or 34) of simulated speech as encoded spatial patterns representing speech phonemes and the characteristic mouth formations that produce the phonemes. Spatial patterns (Fig. 3) may be presented in either tactile or visual form, or both, from the output of a microcomputer speech analyzer (22) that analyzes speech in real time, from a keyboard (50) that generates phoneme- and mouth form- representing signals, or from a memory device (44) that reproduces pre-recorded spatial patterns. The speech analyzer may be incorporated into an armband (90) with a pair of tactile stimulator matrices (36, 38) to provide an unobtrusive prosthetic device for hearing-handicapped individuals. A modified 16 mm projector (240) records spatial patterns on punched film (260) and projects the patterns onto a display (244) to provide a visual presentation.

Abstract translation: 模拟语音方法和装置提供模拟语音的双矩阵表示（32或34）作为表示语音音素的编码空间模式和产生音素的特征口形。 空间模式（图3）可以以触觉或视觉形式或两者从来自生成音素和嘴的键盘（50）的实时分析语音的微型计算机语音分析器（22）的输出呈现 形式表示信号，或来自再现预先记录的空间图形的存储器件（44）。 语音分析器可以并入到具有一对触觉刺激器矩阵（36,38）的臂带（90）中，以为听力障碍的个体提供不引人注目的假体装置。 改进的16mm投影仪（240）记录穿孔胶片（260）上的空间图案并将图案投影到显示器（244）上以提供视觉呈现。

公开(公告)号：WO1982002109A1

公开(公告)日：1982-06-24

申请号：PCT/FI1981000091

申请日：1981-12-15

Applicant: EUROKA OY ,  LAINE UNTO

Inventor： EUROKA OY

IPC: G10L01/00

CPC classification number: G10L13/04 ,  G10L25/15

Abstract: The invention is associated with speech synthesis and with the producing of speech by electronic methods. The object of the invention is to create a new method e.g. for the modelling of the human speech mechanism's acoustic characteristics, i.e., of speech producing. The acoustic transfer function modelling the sound channel is approximated by subdividing it by mathematical methods into partial transfer functions of simpler spectral structure. Each partial transfer function is separately approximated by realizable rational transfer functions. The last mentioned rational transfer functions are realized, each separately, by means of equivalent electrical filters, which have been interconnected in parallel and/or series in the manner implied by the acoustic transfer function which is to be modelled. The models produced by the method of the invention may also be utilized in speech identification, in the estimation of the parameters of a speech signal and in so-called Vocoder apparatus. The invention is also applicable in electronic music synthesizers.

Abstract translation: 本发明与语音合成以及通过电子方式产生语音有关。 本发明的目的是创建一种新的方法，例如， 用于人类语音机制的声学特性，即语音产生的建模。 通过将数学方法细分为更简单的频谱结构的部分传递函数来近似声音通道的声学传递函数。 每个部分传递函数分别由可实现的有理传递函数近似。 最后提到的合理传递函数通过等效电气滤波器分别实现，这些等效电气滤波器以要被建模的声学传递函数所暗示的方式并联和/或串联。 通过本发明的方法产生的模型也可用于语音识别，语音信号的参数估计和所谓的声码器装置中。 本发明也适用于电子音乐合成器。

公开(公告)号：WO1985004747A1

公开(公告)日：1985-10-24

申请号：PCT/US1984002010

申请日：1984-12-04

Applicant: FIRST BYTE

Inventor： FIRST BYTE ,  JACKS, Richard, P. ,  SPRAGUE, Richard, P.

IPC: G10L01/00

CPC classification number: G10L13/04

Abstract: A high-quality, real-time text-to-speech synthesizer system (Fig. 1) handles an unlimited vocabulary with a minimum of hardware by using a microcomputer-software-compatible time domain methodology which requires a minimum of memory and computational power. The system first compares text words to an exception dictionary (Fig. 2). If the word is not found therein, the system applies standard pronunciation rules to the text word. In either instance, the text word is converted to a phoneme sequence. By the use of look-up tables addressed by pointers contained in a phoneme-and-transition matrix (Fig. 3), the synthesizer translates the sequence of phonemes and transitions therebetween into sequences of small speech segments capable of being expressed in terms of repetitions of variable-length portions of short digitally stored waveforms. In general, unvoiced transitions are produced by a sequence of segments which can be concatenated in forward or reverse order to generate different transitions out of the same segments; while voiced transitions are produced by interpolating adjacent phonemes for additioanl savings. Pitch can be varied for naturalness of sound, and/or for intonation changes derived from key words and/or punctuation in the text, by truncating or extending the waveforms of individual voice periods corresponding to voiced segments.

公开(公告)号：WO1984003983A1

公开(公告)日：1984-10-11

申请号：PCT/US1983000464

申请日：1983-03-28

Applicant: EXXON RESEARCH AND ENGINEERING COMPANY

Inventor： EXXON RESEARCH AND ENGINEERING COMPANY ,  BAKER, James, K. ,  MacALLISTER, Jeffrey, G. ,  KLOVSTAD, John, W. ,  SIDELL, Mark, F. ,  BROWN, Peter, F. ,  GANESAN, Kalyan ,  HATTON, Terence, J. ,  LEE, Chin-Hui ,  ROSS, Steven ,  ROTH, Robert, S.

IPC: G10L01/00

CPC classification number: G10L15/00

Abstract: A speech recognition method and apparatus employ a speech processing circuity (26) for repetitively deriving from a speech imput (100), at a frame repetition rate, a plurality of acoustic parameters. The acoustic parameters represent the speech input signal for a frame time. A plurality of template matching and cost processing circuitries (28, 30) are connected to a system bus (24), along with the speech processing circuity, for determining, or identifying, the speech units in the input speech, by comparing the acoustic parameters with stored template patterns. The apparatus can be expanded by adding more template matching and cost processing circuity to the bus thereby increasing the speech recognition capacity of the apparatus. The speech processing circuity establishes overlapping time durations for generating the acoustic parameters and further employs a sinc-Kaiser smoothing function in combination with a folding technique (113) for providing a discrete Fourier transform (112). The Fourier spectra are transformed using a principal component analysis (122) which optimizes the across class variance. The template matching and cost processing circuitries (28, 30) provide distributed processing, on demand, of the acoustic parameters for generating through a dynamic programming technique the recognition decision. Grammar and word model syntax structures reduce the computational load. Template pattern generation is aided by using a "joker" word to specify the time boundaries of utterances spoken in isolation.

Abstract translation: 语音识别方法和装置采用语音处理电路（26），以语音输入（100）以帧重复率重复地导出多个声学参数。 声学参数表示帧时间的语音输入信号。 多个模板匹配和成本处理电路（28,30）连同语音处理电路连接到系统总线（24），用于通过比较声学参数来确定或识别输入语音中的语音单元 具有存储的模板模式。 可以通过向总线添加更多的模板匹配和成本处理电路来扩展该装置，从而增加装置的语音识别能力。 语音处理电路建立用于产生声学参数的重叠时间持续时间，并且还结合用于提供离散付里叶变换（112）的折叠技术（113）来采用sinc-Kaiser平滑函数。 傅立叶光谱使用主成分分析（122）进行变换，该分析优化了跨类别方差。 模板匹配和成本处理电路（28,30）根据需要提供分布式处理声学参数，用于通过动态规划技术生成识别决策。 语法和单词模型语法结构降低了计算量。 通过使用“小丑”字来指定孤立地说出的话语的时间边界来辅助模板模式生成。

公开(公告)号：WO1983003917A1

公开(公告)日：1983-11-10

申请号：PCT/US1982000556

申请日：1982-04-29

Applicant: MASSACHUSETTS INSTITUTE OF TECHNOLOGY ,  FELDMAN, Joel, A. ,  HOFSTETTER, Edward, M.

Inventor： MASSACHUSETTS INSTITUTE OF TECHNOLOGY

IPC: G10L01/00

CPC classification number: G10L19/06

Abstract: A very small, very flexible, high-quality, linear predictive vocoder has been implemented with commercially available integrated circuit. This fully digital realization is based on a distributed signal processing architecture employing three commercial Signal Processing Interface (SPI) single chip microcomputers. One SPI implements a linear predictive speech analyzer (18), a second implements a pitch analyzer (20), while the third implements the excitation generator and synthesizer (28).

Abstract translation: 一个非常小，非常灵活，高质量的线性预测声码器已经实现了商用集成电路。 这种全数字实现基于采用三个商业信号处理接口（SPI）单芯片微型计算机的分布式信号处理架构。 一个SPI实现线性预测语音分析器（18），第二个实现音调分析器（20），而第三个实现激励发生器和合成器（28）。

公开(公告)号：WO1980001014A1

公开(公告)日：1980-05-15

申请号：PCT/US1979000869

申请日：1979-10-22

Applicant: WESTERN ELECTRIC CO INC

Inventor： WESTERN ELECTRIC CO INC ,  PIRZ F ,  RABINER L

IPC: G10L01/00

CPC classification number: G10L25/87 ,  G10L15/063

Abstract: A speech analyzer for recognizing an unknown utterance as one of a set of reference words is adapted (by 118, 119, 103) to generate (105) a feature signal set for each utterance of every reference word. At least one template signal is produced for each reference word which template signal (in 116) is representative of a group of feature signal sets. Responsive to a feature signal set formed (by 105) from the unknown utterance and each reference word template signal, a signal representative of the similarity between the unknown utterance and the template signal is generated (122). A plurality of similarity signals for each reference word is selected and a signal corresponding to the average of said selected similarity signals is formed (135). The average similarity signals are compared to identify the unknown utterance as the most similar reference word (145). Features of the invention include, template formation by successive clustering involving partitioning feature signal sets into groups of predetermined similarity by center point clustering, and recognition by comparing the average of selected similarity measures of a time-warped unknown feature signal set with the cluster-derived reference templates for each vocabulary word.

公开(公告)号：WO1985000686A1

公开(公告)日：1985-02-14

申请号：PCT/US1984001177

申请日：1984-07-23

Applicant: ADVANCED MICRO DEVICES, INC.

Inventor： ADVANCED MICRO DEVICES, INC. ,  WANG, Bu-Chin ,  NARAYAN, Shankar, S.

IPC: G10L01/00

CPC classification number: H04B1/667 ,  G10L25/27

Abstract: Apparatus for coding (Fig. 2A) an original speech signal having a waveform, including a waveform coder (14) operative at a low bit rate, for waveform coding the original speech signal to produce a coded signal having distortion, and an adaptive spectral shaping filter (80, 84) for filtering the distortion in the speech signal. The waveform coder has waveform coding data and the filter has filter coefficient data that are used by a decoding apparatus (Fig. 2B) to reconstruct the original speech signal. Also disclosed are various embodiments of speech analyzers and speech synthesizers which are implemented based on the coding and decoding principles of the coding decoding apparatus.

Abstract translation: 用于编码（图2A）具有波形的原始语音信号的装置，包括以低比特率工作的波形编码器（14），用于对原始语音信号进行波形编码，以产生具有失真的编码信号，以及自适应频谱整形 滤波器（80,84），用于滤除语音信号中的失真。 波形编码器具有波形编码数据，并且滤波器具有由解码装置（图2B）使用的滤波器系数数据，以重构原始语音信号。 还公开了基于编码解码装置的编码和解码原理实现的语音分析器和语音合成器的各种实施例。

公开(公告)号：WO1984002793A1

公开(公告)日：1984-07-19

申请号：PCT/US1983002030

申请日：1983-12-23

Applicant: HENRICKSON, Larry, Keith ,  HUNTINGTON, Dorothy, Allison

IPC: G10L01/00

CPC classification number: H03G7/002 ,  H03G9/025 ,  H04R25/356 ,  H04R2225/43

Abstract: A method of and apparatus for processing audio signals in which a measure of amplitude of audio signals in a selected time period is obtained. The audio signals (Fig. 1) for the selected time period are delayed (18) until the measure of amplitude (16) is obtained, and then the delayed audio signals are normalized (20) using the measure of amplitude. High frequency emphasis (14) may be employed prior to obtaining the measure of amplitude. Alternatively, a multi-channel system (Fig. 3) can be employed for processing audio signals in limited frequency bands (32, 34, 36). The method and apparatus are applicable in a variety of applications including hearing aids, audio storage media, broadcast and public address systems, and voice communications such as telephone systems.

Abstract translation: 一种用于处理音频信号的方法和装置，其中获得在选定时间段内音频信号幅度的测量。 所选择的时间周期的音频信号（图1）被延迟（18），直到获得幅度（16）的测量值，然后使用幅度测量来对延迟的音频信号进行归一化（20）。 在获得幅度测量之前可以采用高频重点（14）。 或者，可以采用多通道系统（图3）来处理有限频带（32,34,36）中的音频信号。 该方法和装置可应用于各种应用，包括助听器，音频存储介质，广播和公共广播系统以及诸如电话系统的语音通信。

公开(公告)号：WO1983001526A1

公开(公告)日：1983-04-28

申请号：PCT/SE1982000339

申请日：1982-10-20

Applicant: KÖHLER, Hans, Olof ,  STRÖMVALL, Sten, Kenneth

IPC: G10L01/00

CPC classification number: G10L15/00

Abstract: A method and apparatus for determining concordance between an analysis signal with at least one reference signal. The smoothed signals or envelopes are formed of a signal which is to be examined and at least one reference signal. A pulse train is generated for each of the smoothed signals or envelopes, and comprises pulses present during the time for a predetermined polarity of the smoothed signal or envelopes relative a threshold signal. The pulse trains are compared at regular intervals with each other, simultaneous coincidence of pulses in both trains at a predetermined number of consecutive comparisons constituting the criterion that the signals which are to be examined are in concordance with the reference signal. For carrying out this signal examination there are means (2, 4) adapted for forming said smoothed signals or envelopes of the signals. The outputs from said means are connected to one input of a comparator (6) for each signal. A threshold signal (URef1, URef2) is applied to the second input of the comparators for generating an output pulse train comprising pulses generated at a predetermined polarity of said smoothed signals or envelopes relative the threshold signals. Means (12, 14, 16, 18) are further adapted for comparing the pulse trains and registering simultaneous coincidence of pulses in the two trains for a predetermined number of consecutive comparisons as the criterion that the examined signal and a given reference signal are in concordance. The invention is primarily intended for utilizing in automatic debiting of utilized video service in a hotel or the like. The invention may also be used for monitoring processes or machines which have specific sound spectra for faults.

Abstract translation: 一种用于确定具有至少一个参考信号的分析信号之间的一致性的方法和装置。 平滑的信号或信封由待检查的信号和至少一个参考信号形成。 为每个平滑的信号或包络生成脉冲串，并且包括在平滑信号的预定极性的时间期间存在的脉冲或相对于阈值信号的信封。 脉冲序列以规则的间隔相互比较，两个列中的脉冲在预定数量的连续比较中同时重合，构成要被检查的信号与参考信号一致的标准。 为了执行该信号检查，存在用于形成信号的所述平滑信号或信封的装置（2,4）。 所述装置的输出连接到每个信号的比较器（6）的一个输入端。 阈值信号（URef1，URef2）被施加到比较器的第二输入端，用于产生输出脉冲序列，该输出脉冲串包括相对于阈值信号以所述平滑信号或包络的预定极性产生的脉冲。 装置（12,14,16,18）进一步适用于比较脉冲序列并将预定数量的连续比较的两列中的脉冲同时重合，作为检查信号和给定参考信号一致的标准 。 本发明主要用于在酒店等中自动借记利用的视频服务。 本发明还可以用于具有用于故障的特定声谱的监视过程或机器。

公开(公告)号：WO1982004493A1

公开(公告)日：1982-12-23

申请号：PCT/JP1982000233

申请日：1982-06-18

Applicant: SANYO ELECTRIC CO ,  SUGIURA YOUJI

Inventor： SANYO ELECTRIC CO

IPC: G10L01/00

CPC classification number: G10L13/02

Abstract: Un synthetiseur vocal servant au montage et a la synthese de segments d'elements sonores extraits d'une forme d'onde vocale analogique, qui convertit un signal vocal analogique en un signal numerique, decale relativement les donnees a proximite de l'extremite posterieure du segment d'element sonore precedent et les donnees a proximite de l'extremite du segment d'element sonore suivant au moyen d'un organe de commande arithmetique servant a calculer le degre d'analogie et extrait de maniere cadencee de la memoire les donnees concernant le segment d'element sonore suivant de maniere que ce segment d'element sonore suivant soit relie de la maniere la plus continue au segment d'element sonore precedent. Par consequent, la variation brusque dans la forme d'onde produite au connecteur entre le segment d'element sonore precedent et le segment d'element sonore suivant, c'est-a-dire, le bruit a haute frequence base sur la discontinuite de la forme d'onde, la deterioration du rapport signal/bruit du son synthetise et la deterioration de l'articulation peuvent etre pratiquement eliminees, et l'on peut obtenir un son synthetise ne presentant pas de forme d'onde discontinue ni de variation de la frequence du son au connecteur.