专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

31. 发明专利

JP2972926B2 失效
公开(公告)号：JP2972926B2
公开(公告)日：1999-11-08
申请号：JP23021687
申请日：1987-09-14
申请人： SEIKOO EPUSON KK , HADOSON KK
发明人： TERAJIMA YOSHUKI , FURUYA YASUSHIGE
IPC分类号： G10H1/043 , G10H7/00 , G10L11/00 , G10L3/00

32. 发明专利

JP2972459B2 失效
公开(公告)号：JP2972459B2
公开(公告)日：1999-11-08
申请号：JP26334992
申请日：1992-10-01
申请人： NIPPON DENKI KK
发明人： NAGANO KEIKO
IPC分类号： G10L11/00 , G10L3/00

33. 发明专利

JPH11308309A PORTABLE TERMINAL 失效
公开(公告)号：JPH11308309A
公开(公告)日：1999-11-05
申请号：JP12818898
申请日：1998-04-22
申请人： CANON KK
发明人： SUGAWARA MASAHITO
IPC分类号： G10L15/00 , G10L13/00 , H04B7/26 , H04M1/00 , H04M1/27 , H04M1/82 , H04M11/00 , G10L3/00
摘要： PROBLEM TO BE SOLVED: To provide a portable terminal where the user can properly enter an operation command of the terminal through voice entry even during a speech without giving a sense of displeasure to a speech opposite party. SOLUTION: In a state of awaiting entry of a command, the portable terminal discriminates whether the command entry is made by a key operation or a voice input depending on a setting state of a key switch 110. When the voice command entry has been selected, whether or not a speech through a radio channel is made is discriminated, and when no speech is made, a command is inputted by speech recognition. When a speech is made, a synthesized voice is sent through the radio channel to deliver a message of holding for the purpose of command entry to the opposite party. The portable terminal enters command by speech recognition. Thus, it is prevented that the voice command entry is heard by the speech opposite party so as to avoid a sense of displeasure form giving to the opposite party.

34. 发明专利

JPH11305788A STATIONARY SECTION DETERMINING METHOD AND SEGMENTATION DEVICE FOR VOICED SOUND 失效
公开(公告)号：JPH11305788A
公开(公告)日：1999-11-05
申请号：JP10729198
申请日：1998-04-17
申请人： MATSUSHITA ELECTRIC IND CO LTD
发明人： MOCHIZUKI AKIRA , NISHIMURA HIROFUMI , MINOWA TOSHIMITSU , ARAI YASUHIKO
IPC分类号： G10L15/04 , G10L15/02 , G10L25/93 , G10L3/00 , G10L9/00
摘要： PROBLEM TO BE SOLVED: To determine stationary sections in detail by pitch marks. SOLUTION: A pitch mark is determined as to a voiced sound by discriminating whether or not an inputted voice is voiced or voiceless. Then stationary/ nonstationary determinations are made by pitch mark positions and then a similarity between a pitch waveform cut out of a pitch mark position where a determination is made on the basis of the determination result and pitch waveforms cut out from pitch mark position before and after it is used. When this similarity exceeds a certain reference value, it is judged that the pitch mark position where the determination is made is in a stationary section. Consequently, feature quantities such as a spectrum are found by pitch waveforms, so the feature quantities are not smoothed; and stationary/nonstationary determination is made by the pitch mark positions, so stationary sections can be determined with precision finer than that of a frame analysis.

35. 发明专利

JPH11305787A VOICE SYNTHESIZING DEVICE 失效
公开(公告)号：JPH11305787A
公开(公告)日：1999-11-05
申请号：JP11208498
申请日：1998-04-22
申请人： VICTOR COMPANY OF JAPAN
发明人： WADA YUJI
IPC分类号： G10L13/08 , G06F17/30 , G10L13/06 , G10L3/00 , G10L5/04
摘要： PROBLEM TO BE SOLVED: To synthesize a natural voice without giving a sense of incongruity through a simple process. SOLUTION: A text data processing part 2 divides text data from a text data input part 1 into divided text data according to specified sections and performs retrieval to check whether or not the divided text data are registered previously in data bases provided in the specified sections. According to the above retrieval, a waveform data selection part 3 selects optimum waveform data out of data stored in respective waveform data storage parts provided in the sections. A waveform data connection part 4 connects the selected waveform data to generate a waveform data sequence.

36. 发明专利

JP2969527B2 失效
公开(公告)号：JP2969527B2
公开(公告)日：1999-11-02
申请号：JP7548590
申请日：1990-03-27
申请人： NITSUTSUKO KK
发明人： IGARASHI MOTOMASA
IPC分类号： G10L11/00 , G10H1/00 , G10K15/04 , G10L15/10 , G10L17/00 , G10L3/00

37. 发明专利

JP2968976B2 失效
公开(公告)号：JP2968976B2
公开(公告)日：1999-11-02
申请号：JP9000190
申请日：1990-04-04
申请人： SATO KUNIO , SHIN KYATAPIRAA MITSUBISHI KK
发明人： SATO KUNIO , OONISHI KEIJIRO
IPC分类号： G06F15/18 , G06N3/00 , G06N99/00 , G10L15/02 , G10L15/08 , G10L15/10 , G10L15/16 , G10L3/00 , G10L5/06 , G10L7/08

38. 发明专利

JPH11298602A METHOD AND SYSTEM FOR TESTING VOICE RESPONCE DEVICE 失效
公开(公告)号：JPH11298602A
公开(公告)日：1999-10-29
申请号：JP10297198
申请日：1998-04-14
申请人： NEC CORP
发明人： ODA TOSHIO
IPC分类号： G10L13/00 , H04M3/22 , G10L3/00
摘要： PROBLEM TO BE SOLVED: To efficiently test a voice responce system provided with plural voice responce devices. SOLUTION: An exchange control part 106 instructs the transmission of voice while designating the arbitrary kind of the voice, an editing method and an idle time slot to plural voice responding devices 101-1 to 101-4 and instructs an exchange 105 to multiplex the said time slot on a highway connected with a testing device 107. The testing device 107 drops the input data of plural time slots through a separating circuit 110 and stores them in a voice memory 108 provided in the testing device 107. By detecting the soundless area in the stored voice data, a test control part 109 detects the head of the voice and by comparing the plural pieces of dropped data, the normal operation of the voice responding devices 101-1 to 101-4 is confirmed.

39. 发明专利

JPH11296515A LANGUAGE MODEL APPROXIMATION LEARNING DEVICE, ITS METHOD AND STORAGE MEDIUM RECORDING APPROXIMATION LEARNING PROGRAM 失效
公开(公告)号：JPH11296515A
公开(公告)日：1999-10-29
申请号：JP9948898
申请日：1998-04-10
申请人： NIPPON TELEGRAPH & TELEPHONE
发明人： MAEDA YASUNARI
IPC分类号： G06F15/18 , G06F17/27 , G06N3/00 , G10L15/06 , G10L15/18 , G10L15/28 , G10L3/00
摘要： PROBLEM TO BE SOLVED: To express by parameters of a number smaller than a true model and to learn an approximate model approximate to the true model by outputting a low-order n-gram approximate model approximate to a true n-gram model from learning data with Kullback Leibler(KL) information quantity as an evalua tion scale. SOLUTION: In the approximation learning device of a language model expressing by a parameter, an n-gram Bayesian learning means 100 receives a word group as learning data to calculate a Bayesian estimated value concerning KL information quantity with a true n-gram model corresponding to the language model to output an n-gram Bayesian estimated model. A low-order n-gram learning means 200 receives the n-gram Bayesian estimated model learned by the means 100 and calculates the low-order n-gram approximation model expressed by the number of parameters smaller than the n-gram Bayesian estimated model with the KL information quantity as the evaluating scale.

40. 发明专利

JPH11296340A INFORMATION OUTPUTTING DEVICE AND METHOD THEREFOR AND RECORDING MEDIUM FOR RECORDING INFORMATION OUTPUT PROGRAM 失效
公开(公告)号：JPH11296340A
公开(公告)日：1999-10-29
申请号：JP9637198
申请日：1998-04-08
申请人： NIPPON TELEGRAPH & TELEPHONE
发明人： FUJIOKA KENGO , NAGAI YOSHIFUMI
IPC分类号： G06F3/16 , G10L13/00 , G10L13/04 , G10L21/06 , G10L3/00 , G10L5/02
摘要： PROBLEM TO BE SOLVED: To output information in sound by making the information easy to listen to by adding information whose listening difficulty is eased to information which is difficult to listen to. SOLUTION: When an information demand from a user is received by an information demand inputting part 1, information desired by the user is retrieved from desired information DB(data base) 3 by a DB retrieving part 2, and the retrieved desired information is outputted in a voice from a voice outputting part 4. Then, a demand for re-listening to the voice output information from a user is accepted by a re-listening demand inputting part 5, and explanation information for easing the listening difficulty of the information to which the user demands to re-listen is obtained from an explanation information DB 7 by an explanation information obtaining part 6, and the obtained explanation information is re-outputted with the information to which the user demands to re-listen with a voice to the user.

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