公开(公告)号：WO99040568A1

公开(公告)日：1999-08-12

申请号：PCT/DE1998/003812

申请日：1998-12-30

IPC: G10L15/07 ,  G10L15/16 ,  G10L19/00 ,  G10L5/00

CPC classification number: G10L15/07 ,  G10L15/16 ,  G10L19/0018

Abstract: In order to transmit voice data, the voice data flow is broken down into phonemes. A code character is assigned to each phoneme in a selective language and/or speaker-specific phoneme catalog (PN1, PN2) and transmitted to a voice synthesis device (SS) located at the transmission target (SD2), whereby the amount of data to be transmitted is considerably reduced. The voice data flow is broken down into phonemes by a neuronal network (NN), which is trained to recognize the phonemes stored in the selective language and/or speaker-specific phoneme catalog (PN1, PN2). The flow of code characters received is once again converted into a sequence of phonemes and emitted by the voice synthesis device (SS).

Abstract translation: 用于语音数据流的语音数据的传输被划分成音素和以可选择的语音和/或说话者特定音素目录（PN1，PN2）为每个音素相关联的代码符号在发送目的地（SD2）语音合成器（SS）发送此，从而将被发送 数据IA的量 大大降低。 语音数据流的分解成音素是由经过训练的音素被存储在识别所选择的语言和/或说话者特定音素目录（PN1，PN2）的神经网络（NN）中进行。 从语音合成装置（SS）接收到的代码符号流被转换回成音素和输出的序列。

公开(公告)号：WO9745830A3

公开(公告)日：1998-02-05

申请号：PCT/IB9700545

申请日：1997-05-13

Applicant: PHILIPS ELECTRONICS NV ,  PHILIPS NORDEN AB

Inventor： VELDHUIS RAYMOND NICOLAAS JOHA ,  KAUFHOLZ PAUL AUGUSTINUS PETER

IPC: G10L19/00 ,  G10L19/12 ,  G10L5/00

CPC classification number: G10L19/12 ,  G10L19/00

Abstract: For coding human speech for subsequent audio reproduction thereof, a plurality of speech segments is derived from speech received, and systematically stored in a data base for later concatenated readout. After the deriving, respective speech segments are fragmented into temporally consecutive source frames, similar source frames as governed by a predetermined similarity measure thereamongst that is based on an underlying parameter set are joined, and joined source frames are collectively mapped onto a single storage frame. Respective segments are stored as containing sequenced referrals to storage frames for therefrom reconstituting the segment in question.

公开(公告)号：WO99063521A1

公开(公告)日：1999-12-09

申请号：PCT/US1999/012427

申请日：1999-06-03

IPC: G10L19/00 ,  G10L5/00

CPC classification number: G10L19/012

Abstract: A signal including speech and background noise is encoded by first decomposing the signal into speech and noise components. A first speech encoding algorithm is then used to generate codebook indices for the speech component and a second algorithm is applied to generate codebook indices for the noise component. The speech encoding algorithm performs better since it faces clean speech, while a simple, very low bit rate algorithm may be used to encode the noise.

Abstract translation: 通过首先将信号分解为语音和噪声分量来编码包括语音和背景噪声的信号。 然后使用第一语音编码算法来产生用于语音分量的码本索引，并且应用第二算法来生成噪声分量的码本索引。 语音编码算法表现得更好，因为它面对清晰的语音，而简单的，非常低的比特率算法可用于对噪声进行编码。

公开(公告)号：WO9627183A3

公开(公告)日：1996-10-10

申请号：PCT/FI9600116

申请日：1996-02-28

Applicant: NOKIA TELECOMMUNICATIONS OY ,  VAINIO JANNE

Inventor： VAINIO JANNE

IPC: G10L19/005 ,  H04L1/00 ,  G10L5/00

CPC classification number: H04L1/0072 ,  G10L19/005 ,  H04L1/0061 ,  H04L1/007 ,  H04L2001/0098

Abstract: The present invention relates to processing speech coding parameters in a telecommunication system. The speech coding parameters of a speech frame, produced by a speech encoder, are divided into groups, i.e. so-called virtual channels, in which speech parameter error correction, channel coding and processing of error-free or erroneous speech parameters are performed independently. At the receiving end, the processing (505) of erroneous and error-free speech parameters can thus be controlled independently on each virtual transmission channel (502) according to the quality of each virtual transmission channel. The speech parameters of the high-quality virtual channels of a speech frame can thus be processed as error-free, replacing the speech coding parameters of the low-quality virtual channels only. The independently processed (505) speech parameters of the virtual channels are thus reassembled (507) into a speech frame, which is applied to decoding. Since part of the information of also erroneous speech frames is utilized, the use of speech information received from a transmission channel can be increased in speech decoding, which reduces for instance interruptions occurring in speech as compared with a situation where all speech frames erroneous even to a slight degree were discarded. The increased and more focused error indication also reduces the number of undetected errors and thus reduces significantly the worst audible disturbances.

公开(公告)号：WO2006030289A1

公开(公告)日：2006-03-23

申请号：PCT/IB2005/002724

申请日：2005-09-14

Applicant: DIGITAL RISE TECHNOLOGY CO., LTD.

Inventor： YOU, Yuli

IPC: G10L5/00

CPC classification number: G10L19/025 ,  G10L19/008 ,  G10L19/032

Abstract: A low bit rate digital audio coding system includes an encoder which assigns codebooks to groups of quantization indexes based on their local properties resulting in codebook application ranges that are independent of block quantization boundaries. The invention also incorporates a resolution filter bank, or a tri-mode resolution filter bank, which is selectively switchable between high and low frequency resolution modes or high, low and intermediate modes such as when detecting transient in a frame. The result is a multichannel audio signal having a significantly lower bit rate for efficient transmission or storage. The decoder is essentially an inverse of the structure and methods of the encoder, and results in a reproduced audio signal that cannot be audibly distinguished from the original signal.

Abstract translation: 低比特率数字音频编码系统包括：编码器，其基于其本地属性将码本分配给量化索引组，导致独立于块量化边界的码本应用范围。 本发明还包括分辨率滤波器组或三模式分辨率滤波器组，其可以在高频和低频分辨率模式或高，低和中间模式之间选择性地切换，例如当检测到帧中的瞬态时。 结果是具有用于有效传输或存储的显着较低比特率的多声道音频信号。 解码器基本上是编码器的结构和方法的逆，并且导致不能与原始信号可听区分的再现音频信号。

公开(公告)号：WO98054695A1

公开(公告)日：1998-12-03

申请号：PCT/US1998/009705

申请日：1998-05-14

IPC: H04M3/38 ,  H04M3/493 ,  G10L5/00 ,  G10L9/08 ,  H04M11/00

CPC classification number: G10L17/04 ,  G10L15/08 ,  G10L15/10 ,  G10L17/005 ,  G10L17/24 ,  G10L25/12 ,  G10L25/24 ,  G10L2015/0638 ,  H04M3/38 ,  H04M3/382 ,  H04M3/385 ,  H04M3/493 ,  H04M2201/40

Abstract: A method of accessing a dial-up service involves the following steps: (a) dialing a service number (172); (b) speaking a number of digits to form a first utterance (174); (c) recognizing the digits using speaker independent speaker recognition (176); (d) when a user has used the dial-up service previously, verifying the user based on the first utterance using a speaker verification system (178); (e) when the user cannot be verified, requesting the user enter a personal identification number (182); and (f) when the personal identification number is valid (184), providing access the dial-up service (186).

Abstract translation: 接入拨号服务的方法包括以下步骤：（a）拨打服务号码（172）; （b）说出一些数字形成第一个发音（174）; （c）使用讲话者独立的演讲人识别识别数字（176）; （d）当用户以前使用过拨号服务时，使用扬声器验证系统（178）基于第一话语来验证用户; （e）当用户不能被验证时，请求用户输入个人识别号码（182）; 和（f）当个人识别号码有效时（184），提供接入拨号服务（186）。

公开(公告)号：WO0128222A3

公开(公告)日：2002-02-28

申请号：PCT/IB0001837

申请日：2000-10-11

Applicant: PERCEPTION DIGITAL TECHNOLOGY ,  LAU JACK ,  TSUI CHI YING ,  CHENG ROGER SHU KWAN ,  YUNG CHI WAI ,  TANG JIMMY TAI KWAN ,  NG KIN PING ,  LAI SAI KIT ,  CHAN KAI KIN ,  CHAN WING CHAU

Inventor： LAU JACK ,  TSUI CHI YING ,  CHENG ROGER SHU KWAN ,  YUNG CHI WAI ,  TANG JIMMY TAI KWAN ,  NG KIN PING ,  LAI SAI KIT ,  CHAN KAI KIN ,  CHAN WING CHAU

IPC: G11B20/00 ,  G11B20/10 ,  G11B27/00 ,  G11B27/10 ,  G11B27/34 ,  G11B31/00 ,  H04S1/00 ,  G06F12/02 ,  G06F12/08 ,  G06F17/30 ,  G06F17/60 ,  G06T1/60 ,  G11B19/00 ,  H04N5/93 ,  H04N7/16 ,  G10L3/02 ,  G10L5/00 ,  G10L7/00 ,  G10L9/00

CPC classification number: G11B27/002 ,  G11B20/00007 ,  G11B20/10527 ,  G11B27/105 ,  G11B27/34 ,  G11B31/00 ,  G11B2220/20 ,  G11B2220/213 ,  G11B2220/216 ,  G11B2220/2545 ,  G11B2220/41 ,  H04S1/007

Abstract: A musical jukebox (10) is disclosed which provides for: fast archiving of songs; a flexible user interface; easy and convenient entry of, access to and/or display (22) of data relating to songs archived by the jukebox (10); easy and convenient search and locate capabilities for locating, reviewing, retrieving and/or playing songs stored in the jukebox; and low cost relative to the functionality, features, conveniences and user-friendliness provided by the jukebox (10). Fast of flush archiving of songs (as well as other data and signals) is accomplished by first saving sets of data without compression, which allows the data to be entered quickly, and then, compressing later at an appropriate time. An unique MP3 bit allocation encoding scheme is used to compress data. An unique memory allocation supports fast data archiving. The user interface employs two-way communication (24) between a remote control (28) and the jukebox (10). A searchable song database is structured to enable very fast searching by music category, and also by title and artist. The jukebox (10) is provided with an on-board song track database to automatically identify new songs input to the jukebox (10).

Abstract translation: 公开了一种音乐自动点唱机（10），其提供：快速归档歌曲; 灵活的用户界面; 简单方便地输入，访问和/或显示与由自动存储塔（10）存档的歌曲有关的数据; 方便搜索和查找定位，查看，检索和/或播放存储在自动存储塔中的歌曲的功能; 并且相对于由点播机（10）提供的功能，特征，便利性和用户友好性而言成本低。 歌曲的快速归档（以及其他数据和信号）通过首先保存数据集而不进行压缩来实现，这允许快速输入数据，然后在适当的时间稍后压缩。 使用独特的MP3位分配编码方案来压缩数据。 唯一的内存分配支持快速数据归档。 用户界面在遥控器（28）和点播机（10）之间采用双向通信（24）。 可搜索的歌曲数据库的结构使得能够通过音乐类别以及标题和艺术家进行非常快速的搜索。 自动存储器（10）设置有板载歌曲轨道数据库，以自动识别输入到自动存储塔（10）的新歌曲。

公开(公告)号：WO00008832A1

公开(公告)日：2000-02-17

申请号：PCT/US1999/016860

申请日：1999-07-26

IPC: G06F13/00 ,  G10L13/04 ,  G10L15/26 ,  H04L12/58 ,  H04M3/42 ,  H04M3/50 ,  H04M3/53 ,  H04M7/12 ,  H04M1/64 ,  G10L5/00

CPC classification number: G10L13/00 ,  H04L51/066 ,  H04M3/5307 ,  H04M7/12 ,  H04M2201/40 ,  H04M2201/60

Abstract: A multi-user e-mail reader system (18) allows several users to access their e-mail accounts simultaneously and have the e-mail messages played back with speech synthesis. The user navigates through various functional states of the system (Fig. 4) using either touch-tone keypad commands or optionally voiced commands interpreted by a speech recognizer (60). Users can send reply e-mail messages without the use of a computer, by invoking the system's text processor (68). The text processor operates in conjunction with a keypad-to-ASCII conversion mechanism (64) that allows fully punctuated and properly addressed e-mail messages to be composed from the touch-tone phone. Digital audio sound file attachments may be recorded through the telephone handset and attached to an outgoing e-mail message. A system for storing canned messages (74) allows the user to quickly send pre-composed reply messages, either as stored or after editing using the text processor. The text processor uses a virtual cursor pointer (72) that may be indexed forward and backward at different granularities (78, 80), depending on whether the system is in play mode or record mode (76). The granularity can also be changed by the user.

Abstract translation: 多用户电子邮件阅读器系统（18）允许多个用户同时访问他们的电子邮件帐户，并且通过语音合成播放电子邮件消息。 用户使用语音识别器（60）解释的触摸键盘命令或可选的浊音命令来浏览系统的各种功能状态（图4）。 用户可以通过调用系统的文本处理器（68）来发送回复电子邮件而不使用计算机。 文本处理器与键盘到ASCII转换机制（64）一起操作，其允许由按键式电话组成的完全标点和正确地寻址的电子邮件消息。 数字音频声音文件附件可以通过电话听筒记录并附加到外发电子邮件。 用于存储固定消息（74）的系统允许用户使用文本处理器快速发送预先编写的应答消息，如存储或编辑之后。 根据系统是处于播放模式还是记录模式（76），文本处理器使用虚拟光标指针（72），虚拟光标指针（72）可根据不同粒度（78,80）向前和向后进行索引。 粒度也可以由用户改变。

公开(公告)号：WO99063520A1

公开(公告)日：1999-12-09

申请号：PCT/EP1999/003765

申请日：1999-05-31

IPC: G10L19/00 ,  G10L19/002 ,  G10L19/005 ,  G10L19/10 ,  G10L19/14 ,  G10L19/18 ,  H04L1/00 ,  G10L5/00

CPC classification number: H04L1/0057 ,  G10L19/002 ,  G10L19/005 ,  G10L19/10 ,  G10L19/18 ,  G10L2019/0008

Abstract: In order to mask errors, the invention provides that binary representations of parameter values are precoded on the transmission side by a linear block code before transmission over a faulty channel. In addition, the redundant information which is added in such a way is not used on the reception side for detecting errors within the binary parameter representations, rather it is exploited in the course of a parameter estimation for improving the quality of the estimated parameter values.

Abstract translation: 用于传输之前在有故障的信道参数值的错误隐藏二进制表示，发送端通过一个线性块码预编码，并且将这样添加的冗余信息没有在用于二进制参数表示内错误检测接收端使用，但作为用于提高所述估计的参数值的质量的参数估计的一部分利用。

公开(公告)号：WO99005671A1

公开(公告)日：1999-02-04

申请号：PCT/US1998/015213

申请日：1998-07-23

IPC: G01B9/02 ,  G01B9/025 ,  G06F3/16 ,  G06F17/27 ,  G10L15/22 ,  G10L15/26 ,  G10L5/00

CPC classification number: G06F17/2785 ,  G06F3/16 ,  G06F17/27 ,  G06F17/277 ,  G10L2015/223 ,  G10L2015/228

Abstract: The present invention is a system for controlling graphical user interface by voice commands. The present invention constitutes a means for receiving issued voice commands from a standard voice recognition system (18), a means for monitoring the state of a target application (16), a means for determining active voice commands from the state of the target application (12), a means for determining whether issued voice command is an active voice command, a means for associating each active voice command with a block of script code data (14), a means for issuing the block of script code data associated with the issued voice command to the graphical user interface when the issued voice command is determined to be an active voice command.

Abstract translation: 本发明是一种通过语音命令来控制图形用户界面的系统。 本发明构成了用于从标准语音识别系统（18）接收发出的语音命令的装置，用于监视目标应用程序（16）的状态的装置，用于从目标应用程序的状态确定主动语音命令的装置（ 12），用于确定所发出的语音命令是否是主动语音命令的装置，用于将每个主动语音命令与脚本代码数据块相关联的装置（14），用于发出与发出的语音命令相关联的脚本代码数据块的装置 当发出的语音命令被确定为主动语音命令时，向图形用户界面发出语音命令。