公开(公告)号：EP2851895A2

公开(公告)日：2015-03-25

申请号：EP14197702.5

申请日：2012-06-29

申请人： Google Inc.

发明人： Ciprian, Chelba I. ,  Xu, Peng ,  Pereira, Fernando

IPC分类号： G10L15/18 ,  G10L15/06

CPC分类号： G10L15/187 ,  G10L15/063 ,  G10L15/14 ,  G10L15/34 ,  G10L2015/0631

摘要： Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for recognizing speech using a variable length of context. Speech data and data identifying a candidate transcription for the speech data are received. A phonetic representation for the candidate transcription is accessed. Multiple test sequences are extracted for a particular phone in the phonetic representation. Each of the multiple test sequences includes a different set of contextual phones surrounding the particular phone. Data indicating that an acoustic model includes data corresponding to one or more of the multiple test sequences is received. From among the one or more test sequences, the test sequence that includes the highest number of contextual phones is selected. A score for the candidate transcription is generated based on the data from the acoustic model that corresponds to the selected test sequence.

摘要翻译： 方法，系统和装置，包括在计算机存储介质上编码的计算机程序，用于使用可变长度的上下文识别语音。 接收用于识别语音数据的候选转录的语音数据和数据。 访问候选转录的语音表示。 在语音表示中为特定电话提取多个测试序列。 多个测试序列中的每一个包括围绕特定电话的不同的上下文电话集合。 指示声学模型包括与多个测试序列中的一个或多个对应的数据的数据被接收。 从一个或多个测试序列中，选择包括最多数量的语境电话的测试序列。 基于来自对应于所选择的测试序列的声学模型的数据生成候选转录的得分。

公开(公告)号：EP1976255B1

公开(公告)日：2015-03-18

申请号：EP08103142.9

申请日：2008-03-28

申请人： Intellisist, Inc.

发明人： Odinak, Gilad

IPC分类号： H04M3/51

CPC分类号： G10L15/30 ,  G10L13/08 ,  G10L15/18 ,  G10L15/193 ,  G10L15/32 ,  G10L15/34 ,  H04M3/51 ,  H04M3/5166 ,  H04M3/5183 ,  H04M2201/38 ,  H04M2201/40

公开(公告)号：EP1082719B1

公开(公告)日：2013-07-03

申请号：EP99915255.6

申请日：1999-04-01

申请人： Koninklijke Philips Electronics N.V.

发明人： LUND, Michael ,  WRIGHT, Karl ,  FAN, Wensheng

IPC分类号： G10L15/28

CPC分类号： G10L15/34 ,  G10L15/187

公开(公告)号：EP1976255A3

公开(公告)日：2010-07-07

申请号：EP08103142.9

申请日：2008-03-28

申请人： Intellisist, Inc.

发明人： Odinak, Gilad

IPC分类号： H04M3/51

CPC分类号： G10L15/30 ,  G10L13/08 ,  G10L15/18 ,  G10L15/193 ,  G10L15/32 ,  G10L15/34 ,  H04M3/51 ,  H04M3/5166 ,  H04M3/5183 ,  H04M2201/38 ,  H04M2201/40

摘要： A system (30) and method (80) for providing an automated call center inline architecture is provided. A plurality of grammar references (65) and prompts are maintained on a script engine (31). A call is received through a telephony interface (32). Audio data (39) is collected using the prompts from the script engine (31), which are transmitted to the telephony interface (32) via a message server (34). Distributed speech recognition (88) is performed on a speech server (33). The grammar references (65) are received from the script engine (31) via the message server (34). Speech results (69) are determined by applying the grammar references (65) to the audio data (39). A new grammar (70) is formed from the speech results (69). Speech recognition results (71) are identified by applying the new grammar (70) to the audio data (39). The speech recognition results (71) are received as a display on an agent console (35).

公开(公告)号：EP2154885A4

公开(公告)日：2010-04-28

申请号：EP08706572

申请日：2008-01-28

申请人： HUAWEI TECH CO LTD

发明人： LIU ZHIHUI ,  YUE ZHONGHUI

IPC分类号： G10L15/26 ,  G06F17/30 ,  G09B21/04 ,  H04L29/06 ,  H04N7/15

CPC分类号： H04N7/15 ,  G09B21/00 ,  G09B21/006 ,  G09B21/009 ,  G10L15/26 ,  G10L15/34 ,  G10L2021/065 ,  H04M3/42391 ,  H04N7/147

公开(公告)号：EP1449203B1

公开(公告)日：2009-08-19

申请号：EP02801823.2

申请日：2002-10-22

申请人： Emma Mixed Signal C.V.

发明人： SHEIKHZADEH-NADJAR, Hamid ,  CORNU, Etienne ,  BRENNAN, Robert ,  DESTREZ, Nicolas ,  DUFAUX, Alain

IPC分类号： G10L15/28

CPC分类号： G10L15/34

摘要： Method and system for real-time speech recognition is provided. The speech algorithm runs on a platform having an input-output processor and a plurality of processor units. The processor units operate substantially in parallel or sequentially to perform feature extraction and pattern matching. While the input-output processor creates a frame, the processor units execute the feature extraction and the pattern matching. Shared memory is provided for supporting the parallel operation.

公开(公告)号：EP1976255A2

公开(公告)日：2008-10-01

申请号：EP08103142.9

申请日：2008-03-28

申请人： Intellisist, Inc.

发明人： Odinak, Gilad

IPC分类号： H04M3/51

CPC分类号： G10L15/30 ,  G10L13/08 ,  G10L15/18 ,  G10L15/193 ,  G10L15/32 ,  G10L15/34 ,  H04M3/51 ,  H04M3/5166 ,  H04M3/5183 ,  H04M2201/38 ,  H04M2201/40

摘要： A system (30) and method (80) for providing an automated call center inline architecture is provided. A plurality of grammar references (65) and prompts are maintained on a script engine (31). A call is received through a telephony interface (32). Audio data (39) is collected using the prompts from the script engine (31), which are transmitted to the telephony interface (32) via a message server (34). Distributed speech recognition (88) is performed on a speech server (33). The grammar references (65) are received from the script engine (31) via the message server (34). Speech results (69) are determined by applying the grammar references (65) to the audio data (39). A new grammar (70) is formed from the speech results (69). Speech recognition results (71) are identified by applying the new grammar (70) to the audio data (39). The speech recognition results (71) are received as a display on an agent console (35).

摘要翻译： 提供了一种用于提供自动呼叫中心在线架构的系统（30）和方法（80）。 许多语法参考（65）和提示被保存在脚本引擎（31）上。 通过电话接口接收呼叫（32）。 使用来自脚本引擎（31）的提示，经由消息服务器（34）发送到电话接口（32）来收集音频数据（39）。 分布式语音识别（88）在语音服务器（33）上执行。 语法参考（65）经由消息服务器（34）从脚本引擎（31）接收。 通过将语法参考（65）应用于音频数据（39）来确定语音结果（69）。 从语音结果（69）形成新的语法（70）。 通过将新语法（70）应用于音频数据（39）来识别语音识别结果（71）。 语音识别结果（71）作为代理控制台（35）上的显示器被接收。

公开(公告)号：EP1852847A4

公开(公告)日：2008-05-21

申请号：EP06711592

申请日：2006-01-12

申请人： NEC CORP

发明人： ISHIKAWA SHINYA ,  YAMABANA KIYOSHI

IPC分类号： G10L15/28 ,  G10L15/34

CPC分类号： G10L15/08 ,  G10L15/34

公开(公告)号：EP1497824A4

公开(公告)日：2006-06-14

申请号：EP03714187

申请日：2003-03-19

申请人： MATSUSHITA ELECTRIC IND CO LTD

发明人： RIGAZIO LUCA ,  NGUYEN PATRICK

IPC分类号： G10L15/08 ,  G10L15/10 ,  G10L15/28 ,  G10L15/00

CPC分类号： G10L15/08 ,  G10L15/10 ,  G10L15/285 ,  G10L15/30 ,  G10L15/34

摘要： A method is provided for improving pattern matching in a speech recognition system having a plurality of acoustic models (20). Similarity measures for acoustic feature vectors (54) are determined in groups that are then buffered into cache memory (59). To further reduce computational processing, the acoustic data may be partitioned amongst a plurality of processing nodes (66, 67, 68). In addition, a priori knowledge of the spoken order may be used to establish the access order (124) used to copy records from the main speech parameter table (120, 200) into a sub-table (130, 204). The sub-table is processed such that the entries are in contiguous memory locations (206) and sorted according to the processing order (208). The speech processing algorithm is then directed to operate upon the sub-table (210) which causes the processor to load the sub-table into high speed cache memory (104, 212).

公开(公告)号：EP1082719A1

公开(公告)日：2001-03-14

申请号：EP99915255.6

申请日：1999-04-01

申请人： Koninklijke Philips Electronics N.V.

发明人： LUND, Michael ,  WRIGHT, Karl ,  FAN, Wensheng

IPC分类号： G10L15/26

CPC分类号： G10L15/34 ,  G10L15/187

摘要： A speech recognition approach that involves forming a series of segments associated with a spoken utterance. Each segment has a time interval within the utterance, and scores characterizing the degree of match of the utterance in that time interval with a set of subword units. Based on the series of segments, the approach includes determining a set of word sequences hypotheses associated with the utterance and then computing scores for the set of word sequence hypotheses using a second set of subword units to represent words in the word sequence hypotheses.