-
公开(公告)号:US20240395239A1
公开(公告)日:2024-11-28
申请号:US18795734
申请日:2024-08-06
Applicant: Google LLC
Inventor: Daisy Antonia Stanton , Sean Matthew Shannon , Soroosh Mariooryad , Russell John Wyatt Skerry-Ryan , Eric Dean Battenberg , Thomas Edward Bagby , David Teh-Hwa Kao
IPC: G10L13/08 , G06N3/08 , G10L13/027
Abstract: Systems and methods for text-to-speech with novel speakers can obtain text data and output audio data. The input text data may be input along with one or more speaker preferences. The speaker preferences can include speaker characteristics. The speaker preferences can be processed by a machine-learned model conditioned on a learned prior distribution to determine a speaker embedding. The speaker embedding can then be processed with the text data to generate an output that includes audio data descriptive of the text data spoken by a novel speaker.
-
公开(公告)号:US10593352B2
公开(公告)日:2020-03-17
申请号:US16001140
申请日:2018-06-06
Applicant: Google LLC
Inventor: Gabor Simko , Maria Carolina Parada San Martin , Sean Matthew Shannon
IPC: G10L17/00 , G10L25/78 , G10L15/22 , G10L15/187 , G10L15/065 , G10L15/18
Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for detecting an end of a query are disclosed. In one aspect, a method includes the actions of receiving audio data that corresponds to an utterance spoken by a user. The actions further include applying, to the audio data, an end of query model. The actions further include determining the confidence score that reflects a likelihood that the utterance is a complete utterance. The actions further include comparing the confidence score that reflects the likelihood that the utterance is a complete utterance to a confidence score threshold. The actions further include determining whether the utterance is likely complete or likely incomplete. The actions further include providing, for output, an instruction to (i) maintain a microphone that is receiving the utterance in an active state or (ii) deactivate the microphone that is receiving the utterance.
-
公开(公告)号:US11551709B2
公开(公告)日:2023-01-10
申请号:US16778222
申请日:2020-01-31
Applicant: Google LLC
Inventor: Gabor Simko , Maria Carolina Parada San Martin , Sean Matthew Shannon
IPC: G10L25/78 , G10L15/22 , G10L15/187 , G10L15/065 , G10L15/18
Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for detecting an end of a query are disclosed. In one aspect, a method includes the actions of receiving audio data that corresponds to an utterance spoken by a user. The actions further include applying, to the audio data, an end of query model. The actions further include determining the confidence score that reflects a likelihood that the utterance is a complete utterance. The actions further include comparing the confidence score that reflects the likelihood that the utterance is a complete utterance to a confidence score threshold. The actions further include determining whether the utterance is likely complete or likely incomplete. The actions further include providing, for output, an instruction to (i) maintain a microphone that is receiving the utterance in an active state or (ii) deactivate the microphone that is receiving the utterance.
-
公开(公告)号:US10929754B2
公开(公告)日:2021-02-23
申请号:US16711172
申请日:2019-12-11
Applicant: Google LLC
Inventor: Shuo-yiin Chang , Bo Li , Gabor Simko , Maria Carolina Parada San Martin , Sean Matthew Shannon
Abstract: A method for training an endpointer model includes short-form speech utterances and long-form speech utterances. The method also includes providing a short-form speech utterance as input to a shared neural network, the shared neural network configured to learn shared hidden representations suitable for both voice activity detection (VAD) and end-of-query (EOQ) detection. The method also includes generating, using a VAD classifier, a sequence of predicted VAD labels and determining a VAD loss by comparing the sequence of predicted VAD labels to a corresponding sequence of reference VAD labels. The method also includes, generating, using an EOQ classifier, a sequence of predicted EOQ labels and determining an EOQ loss by comparing the sequence of predicted EOQ labels to a corresponding sequence of reference EOQ labels. The method also includes training, using a cross-entropy criterion, the endpointer model based on the VAD loss and the EOQ loss.
-
公开(公告)号:US10706840B2
公开(公告)日:2020-07-07
申请号:US15846634
申请日:2017-12-19
Applicant: Google LLC
Inventor: Hasim Sak , Sean Matthew Shannon
IPC: G10L15/14 , G10L15/02 , G10L15/183 , G10L15/06 , G10L15/16 , G06N3/02 , G06N3/08 , G10L15/22 , G06N3/04 , G06N7/00
Abstract: Methods, systems, and apparatus for performing speech recognition. In some implementations, acoustic data representing an utterance is obtained. The acoustic data corresponds to time steps in a series of time steps. One or more computers process scores indicative of the acoustic data using a recurrent neural network to generate a sequence of outputs. The sequence of outputs indicates a likely output label from among a predetermined set of output labels. The predetermined set of output labels includes output labels that respectively correspond to different linguistic units and to a placeholder label that does not represent a classification of acoustic data. The recurrent neural network is configured to use an output label indicated for a previous time step to determine an output label for the current time step. The generated sequence of outputs is processed to generate a transcription of the utterance, and the transcription of the utterance is provided.
-
公开(公告)号:US12067969B2
公开(公告)日:2024-08-20
申请号:US18302764
申请日:2023-04-18
Applicant: Google LLC
Inventor: Eric Dean Battenberg , Daisy Stanton , Russell John Wyatt Skerry-Ryan , Soroosh Mariooryad , David Teh-Hwa Kao , Thomas Edward Bagby , Sean Matthew Shannon
IPC: G10L13/00 , G10L13/047 , G10L13/10
CPC classification number: G10L13/047 , G10L13/10
Abstract: A method for estimating an embedding capacity includes receiving, at a deterministic reference encoder, a reference audio signal, and determining a reference embedding corresponding to the reference audio signal, the reference embedding having a corresponding embedding dimensionality. The method also includes measuring a first reconstruction loss as a function of the corresponding embedding dimensionality of the reference embedding and obtaining a variational embedding from a variational posterior. The variational embedding has a corresponding embedding dimensionality and a specified capacity. The method also includes measuring a second reconstruction loss as a function of the corresponding embedding dimensionality of the variational embedding and estimating a capacity of the reference embedding by comparing the first measured reconstruction loss for the reference embedding relative to the second measured reconstruction loss for the variational embedding having the specified capacity.
-
公开(公告)号:US20230206898A1
公开(公告)日:2023-06-29
申请号:US17673417
申请日:2022-02-16
Applicant: Google LLC
Inventor: Daisy Antonia Stanton , Sean Matthew Shannon , Soroosh Mariooryad , Russell John-Wyatt Skerry-Ryan , Eric Dean Battenberg , Thomas Edward Bagby , David Teh-Hwa Kao
IPC: G10L13/08 , G06N3/08 , G10L13/027
CPC classification number: G10L13/086 , G06N3/08 , G10L13/027
Abstract: Systems and methods for text-to-speech with novel speakers can obtain text data and output audio data. The input text data may be input along with one or more speaker preferences. The speaker preferences can include speaker characteristics. The speaker preferences can be processed by a machine-learned model conditioned on a learned prior distribution to determine a speaker embedding. The speaker embedding can then be processed with the text data to generate an output that includes audio data descriptive of the text data spoken by a novel speaker.
-
公开(公告)号:US11676625B2
公开(公告)日:2023-06-13
申请号:US17152918
申请日:2021-01-20
Applicant: Google LLC
Inventor: Shuo-Yiin Chang , Bo Li , Gabor Simko , Maria Carolina Parada San Martin , Sean Matthew Shannon
CPC classification number: G10L25/78 , G06F18/214 , G06N3/045 , G06N3/08 , G06N5/046 , G06N20/20 , G10L15/16
Abstract: A method for training an endpointer model includes short-form speech utterances and long-form speech utterances. The method also includes providing a short-form speech utterance as input to a shared neural network, the shared neural network configured to learn shared hidden representations suitable for both voice activity detection (VAD) and end-of-query (EOQ) detection. The method also includes generating, using a VAD classifier, a sequence of predicted VAD labels and determining a VAD loss by comparing the sequence of predicted VAD labels to a corresponding sequence of reference VAD labels. The method also includes, generating, using an EOQ classifier, a sequence of predicted EOQ labels and determining an EOQ loss by comparing the sequence of predicted EOQ labels to a corresponding sequence of reference EOQ labels. The method also includes training, using a cross-entropy criterion, the endpointer model based on the VAD loss and the EOQ loss.
-
公开(公告)号:US11222621B2
公开(公告)日:2022-01-11
申请号:US16879714
申请日:2020-05-20
Applicant: Google LLC
Inventor: Eric Dean Battenberg , Daisy Stanton , Russell John Wyatt Skerry-Ryan , Soroosh Mariooryad , David Teh-hwa Kao , Thomas Edward Bagby , Sean Matthew Shannon
IPC: G10L15/22 , G06N3/00 , G06N3/08 , G10L13/047 , G10L13/10
Abstract: A method for estimating an embedding capacity includes receiving, at a deterministic reference encoder, a reference audio signal, and determining a reference embedding corresponding to the reference audio signal, the reference embedding having a corresponding embedding dimensionality. The method also includes measuring a first reconstruction loss as a function of the corresponding embedding dimensionality of the reference embedding and obtaining a variational embedding from a variational posterior. The variational embedding has a corresponding embedding dimensionality and a specified capacity. The method also includes measuring a second reconstruction loss as a function of the corresponding embedding dimensionality of the variational embedding and estimating a capacity of the reference embedding by comparing the first measured reconstruction loss for the reference embedding relative to the second measured reconstruction loss for the variational embedding having the specified capacity.
-
公开(公告)号:US20210142174A1
公开(公告)日:2021-05-13
申请号:US17152918
申请日:2021-01-20
Applicant: Google LLC
Inventor: Shuo-yiin Chang , Bo Li , Gabor Simko , Maria Corolina Parada San Martin , Sean Matthew Shannon
Abstract: A method for training an endpointer model includes short-form speech utterances and long-form speech utterances. The method also includes providing a short-form speech utterance as input to a shared neural network, the shared neural network configured to learn shared hidden representations suitable for both voice activity detection (VAD) and end-of-query (EOQ) detection. The method also includes generating, using a VAD classifier, a sequence of predicted VAD labels and determining a VAD loss by comparing the sequence of predicted VAD labels to a corresponding sequence of reference VAD labels. The method also includes, generating, using an EOQ classifier, a sequence of predicted EOQ labels and determining an EOQ loss by comparing the sequence of predicted EOQ labels to a corresponding sequence of reference EOQ labels. The method also includes training, using a cross-entropy criterion, the endpointer model based on the VAD loss and the EOQ loss.
-
-
-
-
-
-
-
-
-