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. 发明申请

WO2023059978A1 DELIBERATION OF STREAMING RNN-TRANSDUCER BY NON-AUTOREGRESSIVE DECODING 审中-公开
公开(公告)号：WO2023059978A1
公开(公告)日：2023-04-13
申请号：PCT/US2022/076584
申请日：2022-09-16
申请人： GOOGLE LLC
发明人： WANG, Weiran , HU, Ke , SAINATH, Tara N
IPC分类号： G10L15/16 , G10L15/02 , G10L15/063 , G10L15/1815
摘要： A method (400) includes receiving an initial alignment (234) for a candidate hypothesis (120a) generated by a transducer decoder model (230) during a first pass. Here, the candidate hypothesis corresponds to a candidate transcription for an utterance (106) and the initial alignment for the candidate hypothesis includes a sequence of output labels. Each output label corresponds to a blank symbol or a hypothesized sub-word unit. The method also include receiving a subsequent sequence of audio encodings characterizing the utterance (222). During an initial refinement step, the method also includes generating a new alignment (324) for a rescored sequence of output labels using a non-autoregressive decoder (300). The non-autoregressive decoder is configured to receive the initial alignment for the candidate hypothesis and the subsequent sequence of audio encodings.

2. 发明申请

WO2021150424A1 DELIBERATION MODEL-BASED TWO-PASS END-TO-END SPEECH RECOGNITION 审中-公开
公开(公告)号：WO2021150424A1
公开(公告)日：2021-07-29
申请号：PCT/US2021/013449
申请日：2021-01-14
申请人： GOOGLE LLC
发明人： HU, Ke , SAINATH, Tara, N. , PANG, Ruoming , PRABHAVALKAR, Rohit, Prakash
IPC分类号： G10L15/16 , G10L15/32 , G06N3/04 , G10L15/02 , G06N3/0445 , G06N3/049 , G06N3/084 , G06N3/088 , G10L15/063 , G10L15/1815 , G10L15/187 , G10L19/0018 , G10L2015/081 , G10L2015/085
摘要： A method of performing speech recognition using a two-pass deliberation architecture includes receiving a first-pass hypothesis and an encoded acoustic frame and encoding the first-pass hypothesis at a hypothesis encoder. The first-pass hypothesis is generated by a recurrent neural network (RNN) decoder model for the encoded acoustic frame. The method also includes generating, using a first attention mechanism attending to the encoded acoustic frame, a first context vector, and generating, using a second attention mechanism attending to the encoded first-pass hypothesis, a second context vector. The method also includes decoding the first context vector and the second context vector at a context vector decoder to form a second-pass hypothesis

3. 发明申请

WO2021262365A1 NATURAL LANGUAGE PROCESSING 审中-公开
公开(公告)号：WO2021262365A1
公开(公告)日：2021-12-30
申请号：PCT/US2021/033877
申请日：2021-05-24
申请人： AMAZON TECHNOLOGIES, INC.
发明人： GAO, Qiaozi , VERMA, Divyanshu Brijmohan , THATTAI, Govindarajan Sundaram , PING, Qing , CHENGOTTUSSERIYIL, Joel Joseph , STOJANOVIC, Ivan Vitomir , NIU, Feiyang , TUR, Gokhan , ALLEN, Charles J
IPC分类号： G10L15/22 , G06F3/16 , G10L15/18 , G06F3/167 , G06F40/30 , G06F40/35 , G10L13/08 , G10L15/1815 , G10L15/1822 , G10L15/26 , G10L2015/223
摘要： Devices and techniques are generally described for learning personalized responses to declarative natural language inputs. In various examples, a first natural language input may be received. The first natural language input may correspond to intent data corresponding to a declarative user input. In some examples, a dialog session may be initiated with the first user. An action intended by the first user for the first natural language input may be determined based at least in part on the dialog session. In various examples, first data representing the action may be stored in association with second data representing a state described by at least a portion of the first natural language input.

4. 发明申请

WO2021202010A1 SYSTEM AND METHOD FOR CONDUCTING AN AUTOMATED INTERVIEW SESSION 审中-公开
公开(公告)号：WO2021202010A1
公开(公告)日：2021-10-07
申请号：PCT/US2021/018761
申请日：2021-02-19
申请人： GENESYS TELECOMMUNICATIONS LABORATORIES, INC.
发明人： ASOKAN, Anantha, Krishnan , GOKULAKANNAN, Balaji , JAYARAMAN, Sabariganesh , JAYARAGHAVENDRAN, Vidhyasimhan
IPC分类号： G06Q10/06 , G06Q10/10 , G06Q10/00 , G06Q10/1053 , G10L15/063 , G10L15/1815 , G10L15/22 , G10L2015/223 , G10L25/63
摘要： A method for conducting an automated interview session between a candidate and an automated chat resource includes analyzing a candidate resume to identify a first personality trait; generating a first resume-based question based on the first personality trait; presenting, by the automated chat resource, the first resume-based question to the candidate during the automated interview session based on the identified first personality trait; generating a first trait score based on a first response received from the candidate, wherein the first trait score is indicative of a relevance of the identified first personality trait; generating a first question-answer score based on the first resume-based question and the first response, wherein the first question-answer score is indicative of a first question-answer relevance and is used to determine whether to present a second resume-based question to the candidate; analyzing a communication transcript to identify a second personality trait; generating a first transcript-based question based on the second personality trait; presenting, by the automated chat resource, the first transcript- based question to the candidate during the automated interview session based on the identified second personality trait; analyzing a second response from the candidate to determine whether the candidate possesses the identified second personality trait; generating a second trait score based on the second response in response to determining whether the candidate possesses the identified second personality trait, wherein the second trait score is indicative of a relevance of the identified second personality trait; and generating a second question-answer score based on the first transcript-based question and the second response, wherein the second question-answer score is indicative of a second question-answer relevance and is used to determine whether to present a second transcript-based question to the candidate.

5. 发明申请

WO2022240411A1 SYSTEM AND METHOD TO DETERMINE OUTCOME PROBABILITY OF AN EVENT BASED ON VIDEOS 审中-公开
公开(公告)号：WO2022240411A1
公开(公告)日：2022-11-17
申请号：PCT/US2021/032172
申请日：2021-05-13
申请人： CLIPR CO.
发明人： CHEN, Humphrey , CHIN, Cindy , SLOMAN, Aaron
IPC分类号： G06Q10/10 , G06N20/00 , G06N7/005 , G06T11/206 , G06V20/41 , G06V20/44 , G06V20/46 , G10L15/1815 , G10L15/22 , G10L2015/088 , G10L25/57
摘要： System and method for determining an outcome probability of an event based on videos are disclosed. The method includes receiving the videos of an event, creating a building block model, extracting one of an audio content, a video content from the videos, analysing extracted content, generating an analysis result, analysing an engagement between speaker and participant of event, generating a data lake comprising a keyword library, computing the outcome probability of the event, enabling the building block model to learn from the data lake and the outcome probability computed and representing the at least one outcome probability in a pre-defined format.

6. 发明申请

WO2021248011A1 SYSTEMS AND METHODS FOR DETECTING VOICE COMMANDS TO GENERATE A PEER-TO-PEER COMMUNICATION LINK 审中-公开
公开(公告)号：WO2021248011A1
公开(公告)日：2021-12-09
申请号：PCT/US2021/035921
申请日：2021-06-04
申请人： SYNTIANT
发明人： HOLLEMAN, Jeremiah, H., III
IPC分类号： G06F15/16 , G06F3/00 , G06F15/00 , G10L15/30 , G10L13/00 , G10L15/083 , G10L15/1815 , G10L15/22 , G10L2015/088 , G10L2015/223 , H04M1/271 , H04M1/27453 , H04M1/72436
摘要： A voice-based peer-to-peer communication system may be used to detect voice commands from users to provide a wireless communication voice connection that allows the users to directly communicate with each other. The system may include a first computing device of a first user communicatively coupled to a second computing device of a second user over the wireless connection. The system may process the detected voice command having a phrase, contact name, and voice message. The phrase may include a wake, answer, or stop phrase. The contact name may be utilized to determine whether that contact name matches an entry within a predetermined contact list of the first user, where the matched contact name may be associated with the second user. Finally, the system may generate audio data based on the processed voice command that is then transmitted to the second computing device of the second user over the wireless connection.

7. 发明申请

WO2021126492A1 MACHINE LEARNING FOR INTERPRETATION OF SUBVOCALIZATIONS 审中-公开
公开(公告)号：WO2021126492A1
公开(公告)日：2021-06-24
申请号：PCT/US2020/062014
申请日：2020-11-24
申请人： GOOGLE LLC
发明人： VOLOVICH, Yaroslav , OZTASKENT, Ant , AGUERA-ARCAS, Blaise
IPC分类号： G10L15/24 , H04R1/10 , G10L15/16 , G10L15/22 , G10L15/1815 , G10L2015/223 , H04R1/08 , H04R1/1016 , H04R1/1041 , H04R1/1091 , H04R2225/43
摘要： Provided is an in-ear device and associated computational support system that leverages machine learning to interpret sensor data descriptive of one or more in-ear phenomena during subvocalization by the user. An electronic device can receive sensor data generated by at least one sensor at least partially positioned within an ear of a user, wherein the sensor data was generated by the at least one sensor concurrently with the user subvocalizing a subvocalized utterance. The electronic device can then process the sensor data with a machine-learned subvocalization interpretation model to generate an interpretation of the subvocalized utterance as an output of the machine-learned subvocalization interpretation model.

8. 发明申请

WO2021247205A1 SENTIMENT AWARE VOICE USER INTERFACE 审中-公开
公开(公告)号：WO2021247205A1
公开(公告)日：2021-12-09
申请号：PCT/US2021/031720
申请日：2021-05-11
申请人： AMAZON TECHNOLOGIES, INC.
发明人： MADWED, Isaac Joseph , NEMER, Julia Kennedy , KIM, Joo-Kyung , STROM, Nikko , SAUNDERS, Steven Mack , PANFILI, Laura Maggia , JENTOFT, Anna Caitlin , LEE, Sungjin , THOMAS, David , KIM, Young-Bum , GANGA, Pablo Cesar , GUO, Chenlei , TANG, Shuting , YAO, Zhenyu
IPC分类号： G10L15/22 , G10L25/63 , G06F3/16 , G06F40/20 , G06F40/30 , G06K9/00 , G06F3/167 , G06K9/00496 , G06K9/6271 , G06K9/6297 , G06V10/82 , G06V40/16 , G10L15/1815 , G10L15/26 , G10L2015/223 , G10L2015/225 , G10L2015/227
摘要： Described herein is a system for responding to a frustrated user with a response determined based on spoken language understanding (SLU) processing of a user input. The system detects user frustration and responds to a repeated user input by confirming an action to be performed or presenting an alternative action, instead of performing the action responsive to the user input. The system also detects poor audio quality of the captured user input, and responds by requesting the user to repeat the user input. The system processes sentiment data and signal quality data to respond to user inputs.

9. 发明申请

WO2021202569A1 BATCHING TECHNIQUES FOR HANDLING UNBALANCED TRAINING DATA FOR A CHATBOT 审中-公开
公开(公告)号：WO2021202569A1
公开(公告)日：2021-10-07
申请号：PCT/US2021/024946
申请日：2021-03-30
申请人： ORACLE INTERNATIONAL CORPORATION
发明人： DUONG, Thanh Long , JOHNSON, Mark Edward , VISHNOI, Vishal , VINNAKOTA, Srinivas , HONG, Yu-Heng , JALALUDDIN, Elias Luqman
IPC分类号： G06F40/30 , G06N20/00 , G10L15/063 , G10L15/1815 , G10L15/197 , G10L15/22
摘要： The present disclosure relates to chatbot systems, and more particularly, to batching techniques for handling unbalanced training data when training a model such that bias is removed from the trained machine learning model when performing inference. In an embodiment, a plurality of raw utterances is obtained. A bias eliminating distribution is determined and a subset of the plurality of raw utterances is batched according to the bias-reducing distribution. The resulting unbiased training data may be input into a prediction model for training the prediction model. The trained prediction model may be obtained and utilized to predict unbiased results from new inputs received by the trained prediction model.

10. 发明申请

WO2021201925A1 SPEECH-CONTROLLED HEALTH MONITORING SYSTEMS AND METHODS 审中-公开
公开(公告)号：WO2021201925A1
公开(公告)日：2021-10-07
申请号：PCT/US2020/063338
申请日：2020-12-04
申请人： UDP LABS, INC.
发明人： SAYADI, Omid , YOUNG, Steven Jay , HEWITT, Carl , LUCKOW, Alan
IPC分类号： A61B5/00 , A61B5/08 , A61B5/11 , A61B7/00 , G10L25/66 , A61B2562/0204 , A61B2562/0219 , A61B2562/0247 , A61B5/0022 , A61B5/0205 , A61B5/024 , A61B5/0816 , A61B5/0823 , A61B5/4818 , A61B5/4842 , A61B5/6887 , A61B5/7203 , A61B5/7267 , A61B5/7405 , A61B5/746 , A61B5/747 , A61B5/749 , G06F21/602 , G10L15/1815 , G10L15/22 , G10L2015/223 , G10L25/78 , G16H40/67
摘要： Systems and methods for speech-controlled or speech-enabled health monitoring of a subject are described. A device includes a substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic signals and force signals with respect to the subject, an audio interface configured to communicate with the subject, and a processor in connection with the plurality of sensors and the audio interface. The processor configured to determine biosignals from one or more of the acoustic signals and the force signals to monitor a subject's health status, and detect presence of speech in the acoustic signals. The audio interface configured to interactively communicate with at least one of the subject or an entity associated with the subject based on at least one of an action needed due to the subject's health status and a verbal command in detected speech.

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