专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明授权

US06751590B1 Method and apparatus for performing pattern-specific maximum likelihood transformations for speaker recognition 有权
标题翻译：用于执行用于说话人识别的模式特异性最大似然变换的方法和装置
公开(公告)号：US06751590B1
公开(公告)日：2004-06-15
申请号：US09592205
申请日：2000-06-13
申请人： Upendra V. Chaudhari , Ramesh Ambat Gopinath , Stephane Herman Maes
发明人： Upendra V. Chaudhari , Ramesh Ambat Gopinath , Stephane Herman Maes
IPC分类号： G10L1700
CPC分类号： G10L17/02 , G10L17/04
摘要： The present invention uses acoustic feature transformations, referred to as pattern-specific maximum likelihood transformations (PSMLT), to model the voice print of speakers in either a text dependent or independent mode. Each transformation maximizes the likelihood, when restricting to diagonal models, of the speaker training data with respect to the resulting voice-print model in the new feature space. Speakers are recognized (i.e., identified, verified or classified) by appropriate comparison of the likelihood of the testing data in each transformed feature space and/or by directly comparing transformation matrices obtained during enrollment and testing. It is to be appreciated that the principle of pattern-specific maximum likelihood transformations can be extended to a large number of pattern matching problems and, in particular, to other biometrics besides speech.
摘要翻译：本发明使用称为模式特定最大似然变换（PSMLT）的声学特征变换来以文本依赖或独立模式对扬声器的语音印刷进行建模。每个变换使得在限制对角线模型时相对于新特征空间中产生的语音印刷模型的扬声器训练数据的可能性最大化。通过对每个变换的特征空间中的测试数据的可能性和/或通过直接比较在注册和测试期间获得的变换矩阵的适当比较来识别（即，识别，验证或分类）扬声器。应当理解，模式特定的最大似然变换的原理可以扩展到大量的模式匹配问题，特别是除了语音之外的其他生物特征。

2. 发明授权

US06539352B1 Subword-based speaker verification with multiple-classifier score fusion weight and threshold adaptation 失效
标题翻译：基于子词的说话人验证与多分类器得分融合权重和阈值适应
公开(公告)号：US06539352B1
公开(公告)日：2003-03-25
申请号：US08976280
申请日：1997-11-21
申请人： Manish Sharma , Xiaoyu Zhang , Richard J. Mammone
发明人： Manish Sharma , Xiaoyu Zhang , Richard J. Mammone
IPC分类号： G10L1700
CPC分类号： G10L17/10 , G10L15/04 , G10L15/07 , G10L15/10 , G10L15/16 , G10L15/1815 , G10L17/04
摘要： The voice print system of the present invention is a subword-based, text-dependent automatic speaker verification system that embodies the capability of user-selectable passwords with no constraints on the choice of vocabulary words or the language. Automatic blind speech segmentation allows speech to be segmented into subword units without any linguistic knowledge of the password. Subword modeling is performed using a multiple classifiers. The system also takes advantage of such concepts as multiple classifier fusion and data resampling to successfully boost the performance. Key word/key phrase spotting is used to optimally locate the password phrase. Numerous adaptation techniques increase the flexibility of the base system, and include: channel adaptation, fusion adaptation, model adaptation and threshold adaptation.
摘要翻译：本发明的语音打印系统是一种基于词语的文本相关的自动说话人验证系统，其体现了用户可选择密码的能力，对词汇词或语言的选择没有限制。自动盲语分割允许语音被分割成子词单元，而没有密码的任何语言知识。使用多个分类器执行子字建模。该系统还利用了诸如多分类器融合和数据重采样这样的概念来成功提高性能。关键词/关键短语发现用于最佳地定位密码短语。许多适应技术增加了基本系统的灵活性，包括：信道适应，融合适应，模型适应和阈值适配。

3. 发明授权

US06349280B1 Method and apparatus for speaker recognition 失效
标题翻译：用于说话者识别的方法和装置
公开(公告)号：US06349280B1
公开(公告)日：2002-02-19
申请号：US09262083
申请日：1999-03-04
申请人： Hiroaki Hattori
发明人： Hiroaki Hattori
IPC分类号： G10L1700
CPC分类号： G10L17/14 , G10L15/10
摘要： A method of recognizing a speaker of an input speech according to the distance between an input speech pattern, obtained by converting the input speech to a feature parameter series, and a reference pattern preliminarily registered as feature parameter series for each speaker is provided. Contents of the input and reference speech patterns is obtained by recognition. An identical section, in which the contents of the input and reference speech patterns are identical is determined. The distance between the input and reference speech patterns in the calculated identical content section is determined. The speaker of the input speech is recognized on the basis of the determined distance.
摘要翻译：提供了一种根据通过将输入语音转换为特征参数序列获得的输入语音模式之间的距离和预先登记为每个说话者的特征参数系列的参考模式来识别输入语音的扬声器的方法。通过识别获得输入和参考语音模式的内容。确定输入和参考语音模式的内容相同的相同部分。确定所计算的相同内容部分中的输入和参考语音模式之间的距离。基于确定的距离来识别输入语音的扬声器。

4. 发明授权

US06338036B1 Confirmation notification by apparatus using audio recognition as to the acceptability of an input sound 有权
标题翻译：使用音频识别装置对输入声音的可接受性的确认通知
公开(公告)号：US06338036B1
公开(公告)日：2002-01-08
申请号：US09379358
申请日：1999-08-23
申请人： Yasunaga Miyazawa
发明人： Yasunaga Miyazawa
IPC分类号： G10L1700
CPC分类号： G10L15/22
摘要： When a sound which is to be recognized is input to a device, this invention briefly informs the user of whether the sound has been input in an appropriate state. A sound inputting part which outputs a sound to be recognized, spoken by a user as a plurality of words forming one group, as digitized sound data, a sound analysis part which analyzes the sound data and calculates the sound power and characteristic data, a sound division detection/determination part which detects an effective sound division based upon the sound power which has been obtained in the sound analysis part and determines whether the sound to be recognized has been input in an appropriate state, based on the size of the sound power and the time length of the effective sound division, a sound recognition processing part in which the sound to be recognized is recognized and processed, and an information outputting part which outputs information which shows that the recognition object sound is appropriate immediately after the inputting of the sound to be recognized, are provided.
摘要翻译：当要识别的声音被输入到设备时，本发明简要地通知用户声音是否已经被输入到适当的状态。声音输入部，其输出要被识别的声音，作为形成一组的多个字作为数字化声音数据，分析声音数据并计算声音功率和特征数据的声音分析部分，声音基于声音分析部分中获得的声功率来检测有效声音分割，并且基于声音的大小来确定是否已经以适当的状态输入要识别的声音，有效声音的时间长度，识别和处理要识别的声音的声音识别处理部分以及在输入声音之后立即输出表示识别对象声音适合的信息的信息输出部分被提供。

5. 发明授权

US06233555B1 Method and apparatus for speaker identification using mixture discriminant analysis to develop speaker models 有权
公开(公告)号：US06233555B1
公开(公告)日：2001-05-15
申请号：US09198579
申请日：1998-11-24
申请人： Sarangarajan Parthasarathy , Aaron E. Rosenberg
发明人： Sarangarajan Parthasarathy , Aaron E. Rosenberg
IPC分类号： G10L1700
CPC分类号： G10L17/04 , G10L17/24
摘要： A speaker identification system is provided that constructs speaker models using a discriminant analysis technique where the data in each class is modeled by Gaussian mixtures. The speaker identification method and apparatus determines the identity of a speaker, as one of a small group, based on a sentence-length password utterance. A speaker's utterance is received and a sequence of a first set of feature vectors are computed based on the received utterance. The first set of feature vectors are then transformed into a second set of feature vectors using transformations specific to a particular segmentation unit, and likelihood scores of the second set of feature vectors are computed using speaker models trained using mixture discriminant analysis. The likelihood scores are then combined to determine an utterance score and the speaker's identity is validated based on the utterance score. The speaker identification method and apparatus also includes training and enrollment phases. In the enrollment phase the speaker's password utterance is received multiple times. A transcription of the password utterance as a sequence of phones is obtained, and the phone string is stored in a database containing phone strings of other speakers in the group. In the training phase, the first set of feature vectors are extracted from each password utterance and the phone boundaries for each phone in the password transcription are obtained using a speaker independent phone recognizer. A mixture model is developed for each phone of a given speaker's password. Then, using the feature vectors from the password utterances of all of the speakers in the group, transformation parameters and transformed models are generated for each phone and speaker, using mixture discriminant analysis.

6. 发明授权

US06223156B1 Speech recognition of caller identifiers using location information 失效
标题翻译：使用位置信息对呼叫者标识符进行语音识别
公开(公告)号：US06223156B1
公开(公告)日：2001-04-24
申请号：US09056172
申请日：1998-04-07
申请人： Randy G. Goldberg , Roy Philip Weber
发明人： Randy G. Goldberg , Roy Philip Weber
IPC分类号： G10L1700
CPC分类号： H04M1/271 , G10L15/24 , H04M1/57
摘要： A speech recognition system recognizes a caller identifier received during a telephone call as a speech signal from a caller. The system generates a plurality of caller identifier choices from the speech signal and receives location information of the caller. The system includes a database on which is stored a plurality of caller identifiers indexed to a plurality of location information. The system queries a database based on the received location information and retrieves one or more caller identifiers from the database. The system then selects the recognized caller identifier from the plurality of caller identifier choices based on the retrieved one or more caller identifiers.
摘要翻译：语音识别系统将在电话呼叫期间接收到的呼叫者标识识别为来自呼叫者的语音信号。该系统从语音信号生成多个呼叫者识别符选择，并接收呼叫者的位置信息。该系统包括数据库，在其上存储有索引到多个位置信息的多个呼叫者识别符。系统根据接收到的位置信息查询数据库，并从数据库中检索一个或多个呼叫者标识符。然后，系统基于所检索的一个或多个呼叫者标识符从多个呼叫者标识符选择中选择识别的主叫方标识符。

7. 发明授权

US06182037B2 Speaker recognition over large population with fast and detailed matches 失效
标题翻译：演讲者对大量人群的认识快速而详细
公开(公告)号：US06182037B2
公开(公告)日：2001-01-30
申请号：US08851982
申请日：1997-05-06
申请人： Stephane Herman Maes
发明人： Stephane Herman Maes
IPC分类号： G10L1700
CPC分类号： G10L17/06
摘要： Fast and detailed match techniques for speaker recognition are combined into a hybrid system in which speakers are associated in groups when potential confusion is detected between a speaker being enrolled and a previously enrolled speaker. Thus the detailed match techniques are invoked only at the potential onset of saturation of the fast match technique while the detailed match is facilitated by limitation of comparisons to the group and the development of speaker-dependent models which principally function to distinguish between members of a group rather than to more fully characterize each speaker. Thus storage and computational requirements are limited and fast and accurate speaker recognition can be extended over populations of speakers which would degrade or saturate fast match systems and degrade performance of detailed match systems.
摘要翻译：用于说话人识别的快速和详细的匹配技术被组合成混合系统，其中当在被登记的扬声器和先前登记的扬声器之间检测到潜在的混乱时，扬声器被组合在一起。因此，详细匹配技术仅在快速匹配技术的潜在的饱和开始时被调用，而通过与组的比较的限制和说明者依赖模型的发展来促进详细匹配，主要用于区分组的成员而不是更充分地表征每个演讲者。因此，存储和计算需求是有限的，并且可以在扬声器的群体上扩展快速和准确的扬声器识别，这将使得快速匹配系统降级或饱和，并降低详细匹配系统的性能。

8. 发明授权

US06754628B1 Speaker recognition using cohort-specific feature transforms 失效
标题翻译：使用队列特征变换的扬声器识别
公开(公告)号：US06754628B1
公开(公告)日：2004-06-22
申请号：US09592838
申请日：2000-06-13
申请人： Upendra V. Chaudhari , Stephane H. Maes , Jiri Navratil
发明人： Upendra V. Chaudhari , Stephane H. Maes , Jiri Navratil
IPC分类号： G10L1700
CPC分类号： G10L17/12
摘要： Methods and apparatus for facilitating speaker recognition, wherein, from target data that is provided relating to a target speaker and background data that is provided relating to at least one background speaker, a set of cohort data is selected from the background data that has at least one proximate characteristic with respect to the target data. The target data and the cohort data are then combined in a manner to produce at least one new cohort model for use in subsequent speaker verification. Similar methods and apparatus are contemplated for non-voice-based applications, such as verification through fingerprints.
摘要翻译：用于促进说话者识别的方法和装置，其中，从与目标说话者相关的目标数据提供的与提供的至少一个背景说话者相关的背景数据，从至少具有至少一个背景说话者的背景数据中选择一组队列数据相对于目标数据的一个接近特征。然后将目标数据和队列数据组合以产生至少一个新的队列模型，以用于后续说话者验证。预期用于非基于语音的应用的类似方法和装置，例如通过指纹的验证。

9. 发明授权

US06741962B2 Speech recognition system and standard pattern preparation system as well as speech recognition method and standard pattern preparation method 失效
标题翻译：语音识别系统和标准模式准备系统以及语音识别方法和标准模式准备方法
公开(公告)号：US06741962B2
公开(公告)日：2004-05-25
申请号：US10093110
申请日：2002-03-07
申请人： Kenichi Iso
发明人： Kenichi Iso
IPC分类号： G10L1700
CPC分类号： G10L21/038 , G10L15/02
摘要： A speech recognition system for recognizing an input voice of a narrow frequency band. The speech recognition system includes: a frequency band converting unit for converting the input voice of the narrow frequency band into a pseudo voice of a wide frequency band which covers an entirety of the narrow frequency band and which is wider than the narrow frequency band.
摘要翻译：一种用于识别窄频带的输入语音的语音识别系统。语音识别系统包括：频带转换单元，用于将窄频带的输入语音转换成宽频带的伪语音，其覆盖整个窄频带并且比窄频带宽。

10. 发明授权

US06718306B1 Speech collating apparatus and speech collating method 失效
标题翻译：语音整理装置和语音整理方法
公开(公告)号：US06718306B1
公开(公告)日：2004-04-06
申请号：US09690669
申请日：2000-10-17
申请人： Katsuhiko Satoh , Tsuneharu Takeda
发明人： Katsuhiko Satoh , Tsuneharu Takeda
IPC分类号： G10L1700
CPC分类号： G10L17/08 , G10L17/00 , G10L21/06
摘要： A speech of a registered speaker input from an input unit is converted by a converting unit to a sound spectrogram “A” and stored. As a speech of a speaker to be identified is input from the input unit and converted to a sound spectrogram “B” by the converting unit, a detecting unit detects a partial image including a plurality of templates placed in the registered speech image A by a placing unit, and each of areas on the unknown speech image B in which maximum correlation coefficients are calculated. Then, a determining unit compares a mutual positional relationship of the plurality of templates with a mutual positional relationship of the respective areas in which the maximum correlation coefficients are detected to determine from the degree of difference therebetween the identity between the registered speech and the unknown speech.
摘要翻译：从输入单元输入的已登记扬声器的语音由转换单元转换成声谱“A”并被存储。当要被识别的扬声器的语音从输入单元输入并由转换单元转换成声谱“B”时，检测单元通过一个检测单元检测包括放置在登记语音图像A中的多个模板的部分图像放置单元以及计算最大相关系数的未知语音图像B上的每个区域。然后，确定单元将多个模板的相互位置关系与其中检测最大相关系数的各个区域的相互位置关系根据登记语音和未知语音之间的身份之间的差异程度来确定。

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