公开(公告)号：US20190069217A1

公开(公告)日：2019-02-28

申请号：US16171666

申请日：2018-10-26

申请人： International Business Machines Corporation

发明人： Lisa Seacat Deluca ,  Jeremy A. Greenberger

IPC分类号： H04W48/02 ,  H04W24/08 ,  H04L29/08 ,  G06F3/16 ,  H04L12/28 ,  G10L15/22

CPC分类号： H04W48/02 ,  G06F3/165 ,  G06F3/167 ,  G10L15/22 ,  G10L25/87 ,  G10L2015/228 ,  H04L12/282 ,  H04L12/2821 ,  H04L65/1059 ,  H04W24/08

摘要： A computer-implemented method includes: monitoring, by a user device, calling activity on the user device; detecting, by the user device and based on the monitoring, that a call has started on the user device; providing, by the user device, a pause instruction to an assistant device based on detecting that the call has started on the user device, causing the assistant device to disable speech response functions; detecting, by the user device and based on the monitoring, that the call has ended on the user device; and providing, by the user device, a resume instruction to the assistant device based on detecting that the call has ended on the user device, causing the assistant device to resume speech response functions.

公开(公告)号：US20180308501A1

公开(公告)日：2018-10-25

申请号：US15493948

申请日：2017-04-21

申请人： aftercode LLC

发明人： Marc Everett Johnson ,  Mitchell Young Coopet

IPC分类号： G10L21/028 ,  G10L17/04 ,  G10L15/19 ,  G10L21/10 ,  G10L17/02 ,  G10L25/87

CPC分类号： G10L21/028 ,  G10L15/19 ,  G10L17/02 ,  G10L17/04 ,  G10L21/10 ,  G10L25/87

摘要： Systems and techniques for multi speaker attribution using personal grammar detection are described herein. A waveform may be obtained including speaking content of a plurality of speakers. The waveform may be separated into a plurality of segments using audio filters. Members of the plurality of segments including non-speaking content may be discarded to create a set of speaker segments. A first speaker segment may be transcribed to generate a first transcript. The first transcript may be evaluated to identify a grammar pattern and a natural language pattern. A speaker profile may be created for a speaker of the plurality of speakers using the grammar pattern. The speaker profile may be attributed to the first speaker segment and the first transcript. The first transcript may be output to a display including an indication of the speaker.

公开(公告)号：US20180190271A1

公开(公告)日：2018-07-05

申请号：US15395694

申请日：2016-12-30

申请人： Google Inc.

发明人： Gaurav Bhaya ,  Robert Stets

IPC分类号： G10L15/18 ,  H04L29/06 ,  H04L29/08 ,  H04B17/309 ,  G10L25/69 ,  G10L25/87 ,  G10L25/90 ,  G06F17/30

CPC分类号： G10L15/1822 ,  G06F16/3329 ,  G06F16/3344 ,  G06F16/90332 ,  G06F17/2705 ,  G10L15/22 ,  G10L25/69 ,  G10L25/87 ,  G10L25/90 ,  G10L2015/088 ,  H04B17/309 ,  H04L65/1069 ,  H04L65/602 ,  H04L65/80 ,  H04L67/20 ,  H04L67/22 ,  H04L67/42 ,  H04M3/2236 ,  H04M3/4931

摘要： A feedback control system for data transmissions in voice activated data packet based computer network environment is provided. A system can receive audio signals detected by a microphone of a device. The system can parse the audio signal to identify trigger keyword and request. The system can select a content item using the trigger keyword or request. The content item can be configured to establish a communication session between the device and a third party device. The system can monitor the communication session to measure a characteristic of the communication session. The system can generate a quality signal based on the measured characteristic.

公开(公告)号：US09990919B2

公开(公告)日：2018-06-05

申请号：US15318158

申请日：2014-06-24

申请人： NUANCE COMMUNICATIONS, INC.

发明人： Alfred Dielmann ,  Olivier Divay ,  Maximilian Bisani

IPC分类号： G06F17/21 ,  G10L15/26 ,  G10L15/18 ,  G10L15/19 ,  G10L15/02 ,  G10L15/06 ,  G10L25/87

CPC分类号： G10L15/1822 ,  G06F17/211 ,  G10L15/02 ,  G10L15/063 ,  G10L15/19 ,  G10L15/26 ,  G10L25/87

摘要： Methods and apparatus for speech recognition on user dictated words to generate a dictation and using a discriminative statistical model derived from a deterministic formatting grammar module and user formatted documents to extract features and estimate scores from the formatting graph. The processed dictation can be output as formatted text based on a formatting selection to provide an integrated stochastic and deterministic formatting of the dictation.

公开(公告)号：US20180047389A1

公开(公告)日：2018-02-15

申请号：US15404298

申请日：2017-01-12

申请人： ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTIT INSTITUTE

发明人： Hwa Jeon SONG ,  Byung Ok KANG ,  Jeon Gue PARK ,  Yun Keun LEE ,  Hyung Bae JEON ,  Ho Young JUNG

IPC分类号： G10L15/16 ,  G10L25/87 ,  G10L15/14

CPC分类号： G10L15/16 ,  G10L15/142 ,  G10L25/87

摘要： Provided are an apparatus and method for recognizing speech using an attention-based content-dependent (CD) acoustic model. The apparatus includes a predictive deep neural network (DNN) configured to receive input data from an input layer and output predictive values to a buffer of a first output layer, and a context DNN configured to receive a context window from the first output layer and output a final result value.

公开(公告)号：US20180033332A1

公开(公告)日：2018-02-01

申请号：US15658511

申请日：2017-07-25

申请人： David Nelson

发明人： David Nelson

IPC分类号： G09B19/00 ,  G06F3/0481 ,  G10L25/87 ,  G09B5/06 ,  G06T11/20 ,  G10L21/0272 ,  G06F3/0482 ,  G06F3/16

CPC分类号： G09B19/00 ,  G06F3/04817 ,  G06F3/0482 ,  G06F3/165 ,  G06Q10/107 ,  G06Q50/01 ,  G06T11/206 ,  G09B5/06 ,  G09B5/065 ,  G10L21/0272 ,  G10L25/87 ,  H04L12/1831 ,  H04M3/42221 ,  H04M3/56 ,  H04M2201/42

摘要： The present disclosure provides systems and methods for recording, documenting, and visualizing group conversations. More specifically, the present invention relates to systems and methods that allows users to record conversations, document each speaker, visualize the conversation in real time, play back the conversation with visualization for how the conversation progressed from person to person, and compile result statistics on participation levels.

公开(公告)号：US09876913B2

公开(公告)日：2018-01-23

申请号：US15121859

申请日：2015-02-17

申请人： Dolby Laboratories Licensing Corporation

发明人： Richard J. Cartwright ,  Glenn N. Dickins

IPC分类号： H04M3/56 ,  G10L15/08 ,  G10L21/02 ,  G10L25/78 ,  H04W52/02 ,  H04R3/00 ,  G10L25/87

CPC分类号： H04M3/568 ,  G10L15/08 ,  G10L21/02 ,  G10L25/78 ,  G10L25/87 ,  H04M3/563 ,  H04M2201/14 ,  H04R3/005 ,  H04R2420/01 ,  H04W52/0229 ,  Y02D70/23 ,  Y02D70/25

摘要： In an audio conferencing mixing system of the type taking a plurality of audio input streams of input audio information of conference participants, including mixing transition events and outputting a plurality of audio output streams including output audio information, a method of mixing the audio output streams so as to reduce the detectability of the mixing transition events, the method including the steps of (a) determining that a transition event is to occur; (b) determining that a masking trigger is to occur; (c) scheduling the transition event to substantially occur when the masking event occurs. Change blindness mechanism to mask changes in audio conference mix and maintain perceptual continuity.

公开(公告)号：US20180018010A1

公开(公告)日：2018-01-18

申请号：US15695667

申请日：2017-09-05

申请人： SHENZHEN GOODIX TECHNOLOGY CO., LTD.

发明人： Chang LIU ,  Wangwang YANG ,  Haixiang WANG

IPC分类号： G06F1/32 ,  G06F1/16 ,  A61B5/00 ,  G10L25/87 ,  G10L25/21

CPC分类号： G06F1/3228 ,  A61B5/0022 ,  A61B5/0205 ,  A61B5/6801 ,  A61B5/7475 ,  G06F1/163 ,  G06F1/3287 ,  G06F19/00 ,  G10L15/22 ,  G10L25/21 ,  G10L25/78 ,  G10L25/87 ,  G16H40/67

摘要： A power supply manageable wearable device includes: a power management module connected to the microphone interface, a control module and a function module that are connected to the power management module respectively; wherein the control module monitors whether a user instruction includes a voice communication instruction; if the user instruction includes the voice communication instruction, the power management module is enabled to supply power to a microphone and output a first voltage to supply power to the control module; or otherwise, the power management module is enabled to cut off power supply to the microphone and output a second voltage to supply power to the control module and the function module, or output a first voltage to supply power to the control module and output a second voltage to supply power to the function module. The wearable device has the advantages of convenient use and low cost.

公开(公告)号：US20170270201A1

公开(公告)日：2017-09-21

申请号：US15617256

申请日：2017-06-08

申请人： AT&T INTELLECTUAL PROPERTY I, L.P.

发明人： Yeon-Jun KIM ,  I. Dan MELAMED ,  Bernard S. RENGER ,  Steven Neil TISCHER

IPC分类号： G06F17/30 ,  G06F17/00 ,  G10L25/48 ,  G10L15/08 ,  G10L25/90 ,  G10L25/87 ,  G10L25/78

CPC分类号： G06F17/30761 ,  G06F17/00 ,  G10L15/083 ,  G10L25/48 ,  G10L25/78 ,  G10L25/87 ,  G10L25/90

摘要： Apparatuses, systems, methods, and media for filtering a data stream are provided. The data stream is partitioned into a plurality of data stream segments. An acoustic parameter is measured in each of the data stream segments. It is determined whether the acoustic parameter satisfies a first predetermined condition. The first predetermined condition includes a number of variances, in which the acoustic parameter exceeds a predetermined variance threshold, exceeding a predetermined number threshold. An extraneous portion of the data stream is identified in which the first predetermined condition is satisfied. It is determined whether the extraneous portion satisfies a second predetermined condition in the data stream. The extraneous portion is deleted from the data stream to produce a filtered data stream in response to the second predetermined condition being satisfied.

公开(公告)号：US20170263269A1

公开(公告)日：2017-09-14

申请号：US15064441

申请日：2016-03-08

申请人： INTERNATIONAL BUSINESS MACHINES CORPORATION

发明人： Hong-Kwang J. Kuo ,  Lidia L. Mangu ,  Samuel Thomas

IPC分类号： G10L25/87 ,  G10L15/14 ,  G10L25/30 ,  G10L15/22

CPC分类号： G10L25/87 ,  G10L15/142 ,  G10L15/22 ,  G10L15/30 ,  G10L15/32 ,  G10L25/30 ,  G10L25/78 ,  G10L2015/225

摘要： An automatic speech recognition system and a method performed by an automatic speech recognition system are provided. The method includes performing at least two passes of speech activity detection on an acoustic utterance uttered by a speaker. The at least two passes include an initial pass and a subsequent pass. The method further includes estimating at least one of feature statistics and transforms for acoustic feature extraction and acoustic modeling based on an output of an initial pass. The method further includes performing automatic speech recognition using an output of the subsequent pass while bypassing an output of the initial pass to recognize the acoustic utterance.