Business Listing Search

A method of operating a voice-enabled business directory search system includes receiving category-business pairs, each category-business pair including a business category and a specific business, and establishing a data structure having nodes based on the category-business pairs. Each node of the data structure is associated with one or more business categories and a speech recognition language model for recognizing specific businesses associated with the one or more businesses categories. 1) A method of searching a business listing with voice commands over the Internet , the method comprising:receiving, over the Internet, from a user terminal, a query spoken by a user, wherein the query spoken by the user includes a speech utterance representing a category of businesses and a speech utterance representing a geographic location;recognizing the geographic location with a speech recognition engine based on the speech utterance representing the geographic location;recognizing the category of businesses with the speech recognition engine based on the speech utterance representing the category of businesses;searching, with one or more processors, a business listing for businesses within both the recognized category of businesses and the recognized geographic location to select businesses responsive to the query spoken by the user; andsending the user terminal at least some of the responsive businesses.2) The method of claim 1 , comprising selecting claim 1 , from a set of speech recognition language models for recognizing speech claim 1 , a subset of speech recognition language models claim 1 , wherein the subset of speech recognition language models is selected based on the recognized location or the recognized category of businesses.3) The method of claim 2 , wherein the set of speech recognition language models includes N-grams in which a probability of a word in a vocabulary is estimated by counting the occurrences of that word in the context of a last N words.4) ...

Business listing search

A method of operating a voice-enabled business directory search system includes receiving category-business pairs, each category-business pair including a business category and a specific business, and establishing a data structure having nodes based on the category-business pairs. Each node of the data structure is associated with one or more business categories and a speech recognition language model for recognizing specific businesses associated with the one or more businesses categories.

Data driven word pronunciation learning and scoring with crowd sourcing based on the word's phonemes pronunciation scores

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining pronunciations for particular terms. The methods, systems, and apparatus include actions of obtaining audio samples of speech corresponding to a particular term and obtaining candidate pronunciations for the particular term. Further actions include generating, for each candidate pronunciation for the particular term and audio sample of speech corresponding to the particular term, a score reflecting a level of similarity between of the candidate pronunciation and the audio sample, wherein the said score for the particular term is obtained by using a minimum of individual scores of phonemes comprising the term. Additional actions include aggregating the scores for each candidate pronunciation and adding one or more candidate pronunciations for the particular term to a pronunciation lexicon based on the aggregated scores for the candidate pronunciations.

COMPUTING DEVICE WITH REMOTE CONTACT LISTS

In one implementation a computer-implemented method includes generating a group of telephone contacts for a first user, wherein the generating includes identifying a second user as a contact of the first user based upon a determination that the second user has at least a threshold email-based association with the first user; and adding the identified second user to the group of telephone contacts for the first user. The method further includes receiving a first request to connect a first telephone device associated with the first user to a second telephone device associated with the second user. The method also includes identifying a contact identifier of the second telephone device using the generated group of telephone contacts for the first user, and initiating a connection between the first telephone device and the second telephone device using the identified contact identifier. 1. A computer-implemented method comprising: identifying, by the computer system, a second user as a contact of the first user based upon a determination that the second user has at least a threshold email-based association with the first user; and', 'adding, by the computer system, the identified second user to the group of telephone contacts for the first user;, 'generating, by a computer system, a group of telephone contacts for a first user, wherein the generating comprisesreceiving, at the computer system, a first request to connect a first telephone device associated with the first user to a second telephone device associated with the second user;identifying, by the computer system, a contact identifier of the second telephone device using the generated group of telephone contacts for the first user; andinitiating, by the computer system, a connection between the first telephone device and the second telephone device using the identified contact identifier.2. The computer implemented method of claim 1 , wherein the threshold email-based association includes a threshold level of ...

CROSS-LINGUAL INITIALIZATION OF LANGUAGE MODELS

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for initializing language models for automatic speech recognition. In one aspect, a method includes receiving logged speech recognition results from an existing corpus that is specific to a given language and a target context, generating a target corpus by machine-translating the logged speech recognition results from the given language to a different, target language, and estimating a language model that is specific to the different, target language and the same, target context, using the target corpus. 1. A computer-implemented method performed by at least one processor , the method comprising:receiving logged speech recognition results from an existing corpus that is specific to a given language and a target context; machine-translating the logged speech recognition results from the given language to a different, target language; and', 'augmenting an existing, partial target corpus specific for the different, target language and the target context with the machine-translated logged speech recognition results; and, 'generating a target corpus byestimating a language model that is specific to the different, target language and the same, target context, using the target corpus.2. The method of claim 1 , wherein estimating the language model comprises counting each occurrence of each distinctive word or phrase in the target corpus.3. The method of claim 2 , wherein estimating the language model comprises determining a relative frequency of occurrence of each distinctive word or phrase in the target corpus claim 2 , from among all distinctive words or phrases in the target corpus.4. The method of claim 1 , wherein the target context is associated with a particular application or application state claim 1 , operating system claim 1 , geographic location or region claim 1 , or environmental or ambient characteristic.5. The method of claim 1 , wherein the target context is ...

SPEECH RECOGNITION WITH PARALLEL RECOGNITION TASKS

The subject matter of this specification can be embodied in, among other things, a method that includes receiving an audio signal and initiating speech recognition tasks by a plurality of speech recognition systems (SRS's). Each SRS is configured to generate a recognition result specifying possible speech included in the audio signal and a confidence value indicating a confidence in a correctness of the speech result. The method also includes completing a portion of the speech recognition tasks including generating one or more recognition results and one or more confidence values for the one or more recognition results, determining whether the one or more confidence values meets a confidence threshold, aborting a remaining portion of the speech recognition tasks for SRS's that have not generated a recognition result, and outputting a final recognition result based on at least one of the generated one or more speech results. 1. A computer-implemented method comprising:receiving, at a computer system, an audio signal;initiating, by the computer system, a plurality of speech recognition tasks for the audio signal;detecting that a portion of the plurality of speech recognition tasks have completed, wherein a remaining portion of the plurality of speech recognition tasks have not yet completed;obtaining recognition results and confidence values for the portion of the plurality of speech recognition tasks, wherein the recognition results identify one or more candidate representations of the audio signal and the confidence values identify one or more probabilities that the recognition results are correct;generating one or more combined confidence values for the recognition results based on the recognition results and the confidence values for the portion of the plurality of speech;determining, by the computer system, whether at least one of the one or more combined confidence values is greater than or equal to a threshold confidence value; andin response to determining ...

RECOGNIZING SPEECH IN MULTIPLE LANGUAGES

Speech recognition systems may perform the following operations: receiving audio; recognizing the audio using language models for different languages to produce recognition candidates for the audio, where the recognition candidates are associated with corresponding recognition scores; identifying a candidate language for the audio; selecting a recognition candidate based on the recognition scores and the candidate language; and outputting data corresponding to the selected recognition candidate as a recognized version of the audio. 1. A method comprising:receiving audio;recognizing the audio using language models for different languages to produce recognition candidates for the audio, the recognition candidates being associated with corresponding recognition scores;identifying a candidate language for the audio;selecting a recognition candidate based on the recognition scores and the candidate language; andoutputting data corresponding to the selected recognition candidate as a recognized version of the audio.2. The method of claim 1 , wherein identification of the candidate language is performed substantially in parallel with recognition of the audio using the language models for different languages.3. The method of claim 1 , wherein identification of the candidate language for the audio occurs prior to recognition of the audio using the language models for different languages.4. The method of claim 1 , wherein selecting the recognition candidate comprises taking agreement of different language models into account in deciding which recognition candidate to select.5. The method of claim 1 , further comprising:selecting the language models.6. The method of claim 5 , wherein the language models are selected based on input from a user from whom the audio is received.7. The method of claim 5 , wherein selecting the language models comprises:identifying languages associated with previously-received audio; andselecting language models corresponding to the identified ...

Speech recognition models based on location indicia

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing speech recognition using models that are based on where, within a building, a speaker makes an utterance are disclosed. The methods, systems, and apparatus include actions of receiving data corresponding to an utterance, and obtaining location indicia for an area within a building where the utterance was spoken. Further actions include selecting one or more models for speech recognition based on the location indicia, wherein each of the selected one or more models is associated with a weight based on the location indicia. Additionally, the actions include generating a composite model using the selected one or more models and the respective weights of the selected one or more models. And the actions also include generating a transcription of the utterance using the composite model.

Speech Recognition with Parallel Recognition Tasks

The subject matter of this specification can be embodied in, among other things, a method that includes receiving an audio signal and initiating speech recognition tasks by a plurality of speech recognition systems (SRS's). Each SRS is configured to generate a recognition result specifying possible speech included in the audio signal and a confidence value indicating a confidence in a correctness of the speech result. The method also includes completing a portion of the speech recognition tasks including generating one or more recognition results and one or more confidence values for the one or more recognition results, determining whether the one or more confidence values meets a confidence threshold, aborting a remaining portion of the speech recognition tasks for SRS's that have not generated a recognition result, and outputting a final recognition result based on at least one of the generated one or more speech results. 1. A computer-implemented method comprising:receiving, at a computer system, an audio signal;initiating, by the computer system, a plurality of speech recognition tasks for the audio signal;detecting that a portion of the plurality of speech recognition tasks have completed, wherein a remaining portion of the plurality of speech recognition tasks have not yet completed;obtaining recognition results and confidence values for the portion of the plurality of speech recognition tasks, wherein the recognition results identify one or more candidate representations of the audio signal and the confidence values identify one or more probabilities that the recognition results are correct;generating one or more combined confidence values for the recognition results based on the recognition results and the confidence values for the portion of the plurality of speech;determining, by the computer system, whether at least one of the one or more combined confidence values is greater than or equal to a threshold confidence value; andin response to determining ...

Computing Device With Remote Contact Lists

In one implementation a computer-implemented method includes generating a group of telephone contacts for a first user, wherein the generating includes identifying a second user as a contact of the first user based upon a determination that the second user has at least a threshold email-based association with the first user; and adding the identified second user to the group of telephone contacts for the first user. The method further includes receiving a first request to connect a first telephone device associated with the first user to a second telephone device associated with the second user. The method also includes identifying a contact identifier of the second telephone device using the generated group of telephone contacts for the first user, and initiating a connection between the first telephone device and the second telephone device using the identified contact identifier. 1. A computer-implemented method comprising: identifying, by the computer system, a second user as a contact of the first user based upon a determination that the second user has at least a threshold email-based association with the first user; and', 'adding, by the computer system, the identified second user to the group of telephone contacts for the first user;, 'generating, by a computer system, a group of telephone contacts for a first user, wherein the generating comprisesreceiving, at the computer system, a first request to connect a first telephone device associated with the first user to a second telephone device associated with the second user;identifying, by the computer system, a contact identifier of the second telephone device using the generated group of telephone contacts for the first user; andinitiating, by the computer system, a connection between the first telephone device and the second telephone device using the identified contact identifier.2. The computer implemented method of claim 1 , wherein the threshold email-based association includes a threshold level of ...

DATA DRIVEN PRONUNCIATION LEARNING WITH CROWD SOURCING

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining pronunciations for particular terms. The methods, systems, and apparatus include actions of obtaining audio samples of speech corresponding to a particular term and obtaining candidate pronunciations for the particular term. Further actions include generating, for each candidate pronunciation for the particular term and audio sample of speech corresponding to the particular term, a score reflecting a level of similarity between of the candidate pronunciation and the audio sample. Additional actions include aggregating the scores for each candidate pronunciation and adding one or more candidate pronunciations for the particular term to a pronunciation lexicon based on the aggregated scores for the candidate pronunciations. 1. A computer-implemented method comprising:obtaining audio samples of speech corresponding to a particular term;obtaining candidate pronunciations for the particular term;generating, for each candidate pronunciation for the particular term and audio sample of speech corresponding to the particular term, a score reflecting a level of similarity between the candidate pronunciation and the audio sample;aggregating the scores for each candidate pronunciation; andadding one or more candidate pronunciations for the particular term to a pronunciation lexicon based on the aggregated scores for the candidate pronunciations.2. The method of claim 1 , wherein adding one or more candidate pronunciations for the particular term comprises:identifying a candidate pronunciation of the candidate pronunciations with an aggregated score that indicates a closer level of similarity between the candidate pronunciation and the audio samples than levels of similarity between the other candidate pronunciations and the audio samples; andadding the identified candidate expression to the pronunciation lexicon.3. The method of claim 1 , wherein adding one or more ...

COOPERATIVELY TRAINING AND/OR USING SEPARATE INPUT AND SUBSEQUENT CONTENT NEURAL NETWORKS FOR INFORMATION RETRIEVAL

Systems, methods, and computer readable media related to information retrieval. Some implementations are related to training and/or using a relevance model for information retrieval. The relevance model includes an input neural network model and a subsequent content neural network model. The input neural network model and the subsequent content neural network model can be separate, but trained and/or used cooperatively. The input neural network model and the subsequent content neural network model can be “separate” in that separate inputs are applied to the neural network models, and each of the neural network models is used to generate its own feature vector based on its applied input. A comparison of the feature vectors generated based on the separate network models can then be performed, where the comparison indicates relevance of the input applied to the input neural network model to the separate input applied to the subsequent content neural network model. 1. A method implemented by one or more processors , the method comprising:receiving a textual query generated based on user interface input provided by a user via a client device of the user, the textual query comprising multiple words;applying a query representation of the textual query to a trained input neural network model, the query representation being based on two or more of the multiple words;generating one query vector over the trained input neural network model based on applying the query representation to the trained input neural network model; 'determining the relevance value based on a dot product of the query vector to one vector stored in association with the content item, the one vector being stored in association with the content item prior to receiving the query;', 'determining a relevance value that indicates relevance of a content item to the query, wherein determining the relevance value comprisesbased on the relevance value, providing to the client device a result that is based on the ...

TEXT CLASSIFICATION AND TRANSFORMATION BASED ON AUTHOR

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for transforming and classifying text based on analysis of training texts from particular authors. One of the methods includes receiving an input text including one or more words and a requested author; generating a vector stream representing the input text based on an encoder language model and including one or more multi-dimensional vectors associated with associated words of the words of the input text and representing a distribution of contexts in which the associated words occurred in a plurality of training texts; and producing an output text representing a particular transformation of the input text based at least in part on a decoder language model, the generated vector stream, and the requested author. 1. A method performed by a system comprising one or more computers , the method comprising:receiving an input text including one or more words and a name of a requested author;generating a vector stream representing the input text based on an encoder language model, wherein the vector stream includes one or more multi-dimensional vectors each associated with one or more associated words of the words of the input text and representing a distribution of contexts in which the associated words occurred in a plurality of training texts processed by the encoder language model; andproducing an output text representing a particular transformation of the input text based at least in part on a decoder language model, the generated vector stream, and the requested author, wherein the decoder language model stores distributions of words used by particular authors in the plurality of training texts that caused the encoder language model to produce particular vectors representing the words.2. The method of claim 1 , wherein the particular transformation of the input text is a transformation of the input text into text written in the style of the requested author.3. The method of ...

Computing Device with Remote Contact Lists

In one implementation a computer-implemented method includes generating a group of telephone contacts for a first user, wherein the generating includes identifying a second user as a contact of the first user based upon a determination that the second user has at least a threshold email-based association with the first user; and adding the identified second user to the group of telephone contacts for the first user. The method further includes receiving a first request to connect a first telephone device associated with the first user to a second telephone device associated with the second user. The method also includes identifying a contact identifier of the second telephone device using the generated group of telephone contacts for the first user, and initiating a connection between the first telephone device and the second telephone device using the identified contact identifier. 120-. (canceled)21. A computer-implemented method comprising:receiving, at a computer system, a request for a list of telephone contacts for a user;accessing, by the computer system, social network information for the user, wherein the social network information identifies interactions between the user and a plurality of other users on one or more social networks;determining, by the computer system, levels of interactions between the user and the plurality of other users based, at least in part, on the social network information that identifies the interactions between the user and the plurality of other users;selecting, by the computer system, a portion of the plurality of other users based, at least in part, on the determined levels of interaction between the user and the plurality of other users;generating, by the computer system, a group of telephone contacts for the user that includes telephone contact entries that correspond to users from the portion of the plurality of other users; andproviding, by the computer system, the group of telephone contacts for presentation to the user ...

Training encoder model and/or using trained encoder model to determine responsive action(s) for natural language input

Systems, methods, and computer readable media related to: training an encoder model that can be utilized to determine semantic similarity of a natural language textual string to each of one or more additional natural language textual strings (directly and/or indirectly); and/or using a trained encoder model to determine one or more responsive actions to perform in response to a natural language query. The encoder model is a machine learning model, such as a neural network model. In some implementations of training the encoder model, the encoder model is trained as part of a larger network architecture trained based on one or more tasks that are distinct from a “semantic textual similarity” task for which the encoder model can be used.

Identifying substitute pronunciations

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, including selecting terms; obtaining an expected phonetic transcription of an idealized native speaker of a natural language speaking the terms; receiving audio data corresponding to a particular user speaking the terms in the natural language; obtaining, based on the audio data, an actual phonetic transcription of the particular user speaking the terms in the natural language; aligning the expected phonetic transcription of the idealized native speaker of the natural language with the actual phonetic transcription of the particular user; identifying, based on the aligning, a portion of the expected phonetic transcription that is different than a corresponding portion of the actual phonetic transcription; and based on identifying the portion of the expected phonetic transcription, designating the expected phonetic transcription as a substitute pronunciation for the corresponding portion of the actual phonetic transcription.

SELECTIVE TEXT PREDICTION FOR ELECTRONIC MESSAGING

A computing system is described that includes user interface components configured to receive typed user input; and one or more processors. The one or more processors are configured to: receive, by a computing system and at a first time, a first portion of text typed by a user in an electronic message being edited; predict, based on the first portion of text, a first candidate portion of text to follow the first portion of text; output, for display, the predicted first candidate portion of text for optional selection to append to the first portion of text; determine, at a second time that is after the first time, that the electronic message is directed to a sensitive topic; and responsive to determining that the electronic message is directed to a sensitive topic, refrain from outputting subsequent candidate portions of text for optional selection to append to text in the electronic message. 1: A method comprising:receiving, by a computing system and at a first time, a first portion of text of an electronic message being edited;predicting, by the computing system and based on the first portion of text, a first candidate portion of text to follow the first portion of text;outputting, for display, the predicted first candidate portion of text for optional selection to append to the first portion of text;determining, by the computing system and at a second time that is after the first time, that the electronic message is directed to a sensitive topic based on a modification to the electronic message performed between the first time and the second time; andresponsive to determining that the electronic message is directed to a sensitive topic, refraining from outputting subsequent candidate portions of text for optional selection to append to text in the electronic message.2: The method of claim 1 , wherein predicting the first candidate portion of text comprises:predicting, by the computing system and using a machine learning model, one or more candidate portions of ...

COOPERATIVELY TRAINING AND/OR USING SEPARATE INPUT AND SUBSEQUENT CONTENT NEURAL NETWORKS FOR INFORMATION RETRIEVAL

Systems, methods, and computer readable media related to information retrieval. Some implementations are related to training and/or using a relevance model for information retrieval. The relevance model includes an input neural network model and a subsequent content neural network model. The input neural network model and the subsequent content neural network model can be separate, but trained and/or used cooperatively. The input neural network model and the subsequent content neural network model can be “separate” in that separate inputs are applied to the neural network models, and each of the neural network models is used to generate its own feature vector based on its applied input. A comparison of the feature vectors generated based on the separate network models can then be performed, where the comparison indicates relevance of the input applied to the input neural network model to the separate input applied to the subsequent content neural network model. 1. A method implemented by one or more processors , comprising: being a responsive reply to the initial content in the corresponding electronic resource, or', 'occurring positionally subsequent to the initial content in the corresponding electronic resource; and, 'the input representation is a representation of initial content of a corresponding electronic resource, and the subsequent content representation is a representation of subsequent content of the corresponding electronic resource, and wherein the subsequent content is included based on it, 'identifying a plurality of positive training instances that each include an input representation and a subsequent content representation, wherein for each of the positive training instances generating an input vector based on applying the input representation to an input neural network model of the relevance model;', 'generating a subsequent content vector based on applying the subsequent content representation to a subsequent content neural network model of the ...

COOPERATIVELY TRAINING AND/OR USING SEPARATE INPUT AND RESPONSE NEURAL NETWORK MODELS FOR DETERMINING RESPONSE(S) FOR ELECTRONIC COMMUNICATIONS

Systems, methods, and computer readable media related to determining one or more responses to provide that are responsive to an electronic communication that is generated through interaction with a client computing device. For example, determining one or more responses to provide for presentation to a user as suggestions for inclusion in a reply to an electronic communication sent to the user. Some implementations are related to training and/or using separate input and response neural network models for determining responses for electronic communications. The input neural network model and the response neural network model can be separate, but trained and/or used cooperatively.

Speech Recognition with Parallel Recognition Tasks

The subject matter of this specification can be embodied in, among other things, a method that includes receiving an audio signal and initiating speech recognition tasks by a plurality of speech recognition systems (SRS's). Each SRS is configured to generate a recognition result specifying possible speech included in the audio signal and a confidence value indicating a confidence in a correctness of the speech result. The method also includes completing a portion of the speech recognition tasks including generating one or more recognition results and one or more confidence values for the one or more recognition results, determining whether the one or more confidence values meets a confidence threshold, aborting a remaining portion of the speech recognition tasks for SRS's that have not generated a recognition result, and outputting a final recognition result based on at least one of the generated one or more speech results. 1. (canceled)2. A computer-implemented method comprising:providing particular audio data to each automated speech recognizer of a set of automated speech recognizers;before all of the automated speech recognizers have completed processing the particular audio data, determining that a particular automated speech recognizer of the set of automated speech recognizers has completed processing of the particular audio data, and that a confidence value associated with the particular automated speech recognizer processing the particular audio data satisfies a particular confidence value threshold; andin response to determining that a particular automated speech recognizer of the set of automated speech recognizers has completed processing of the particular audio data, and that a confidence value associated with the particular automated speech recognizer processing the particular audio data satisfies a particular confidence value threshold, providing an output of the particular automated speech recognizer, of the set of automated speech recognizers ...

Personal Directory Service

A method of providing navigation directions includes receiving, at a user terminal, a query spoken by a user, wherein the query spoken by the user includes a speech utterance indicating (i) a category of business, (ii) a name of the business, and (iii) a location at which or near which the business is disposed; identifying, by processing hardware, the business based on the speech utterance; and providing navigation directions to the business via the user terminal. 1. A method of providing navigation directions , the method comprising:receiving, at a user terminal, a query spoken by a user, wherein the query spoken by the user includes a speech utterance indicating (i) a category of business, (ii) a name of the business, and (iii) a location at which or near which the business is disposed;identifying, by processing hardware, the business based on the speech utterance;and providing navigation directions to the business via the user terminal.2. The method of claim 1 , wherein the speech utterance includes a name of the geographic location and an indication that the business is located near the geographic location.3. The method of claim 1 , wherein the speech utterance includes a name of the geographic location and an indication that the business is located near the geographic location.4. The method of claim 1 , further comprising:prompting the user to provide a geographical location.5. The method of claim 1 , further comprising claim 1 , after recognizing the identifier of the specific business claim 1 , providing information about the specific business to the user.6. The method of claim 6 , wherein providing information about the specific business comprises providing a phone number of the specific business.7. The method of wherein recognizing the identifier of the specific business comprises biasing a speech recognition module towards one or more language models associated with the type of business.8. The method of claim 1 , further comprising:selecting, from a set of ...

Personal Directory Service

A method of providing a personal directory service includes receiving, over the Internet, from a user terminal, a query spoken by a user, where the query spoken by the user includes a speech utterance representing a category of persons. The method also includes determining a geographic location of the user terminal, recognizing the category of persons with the speech recognition engine based on the speech utterance representing the category of persons a listing of persons within or near the determined geographic location matching the query to select persons responsive to the query spoken by the user, and sending to the user terminal information related to at least some of the responsive persons. 1. A method of providing a personal directory service , the method comprising:receiving, over the Internet, from a user terminal, a query spoken by a user, wherein the query spoken by the user includes a speech utterance representing a category of persons;determining a geographic location of the user terminal;recognizing the category of persons with the speech recognition engine based on the speech utterance representing the category of persons;searching, with one or more processors, a listing of persons within or near the determined geographic location matching the query to select persons responsive to the query spoken by the user; andsending to the user terminal information related to at least some of the responsive persons.2. The method of claim 1 , wherein the query further includes a speech utterance representing a name of a person; the method further comprising:recognizing the name of the person with a speech recognition engine based on the speech utterance representing the name of the person;3. The method of claim 1 , wherein the query further includes a speech utterance representing a geographic location; the method further comprising:recognizing the geographic location with a speech recognition engine based on the speech utterance representing the geographic ...

Speech Recognition with Parallel Recognition Tasks

The subject matter of this specification can be embodied in, among other things, a method that includes receiving an audio signal and initiating speech recognition tasks by a plurality of speech recognition systems (SRS's). Each SRS is configured to generate a recognition result specifying possible speech included in the audio signal and a confidence value indicating a confidence in a correctness of the speech result. The method also includes completing a portion of the speech recognition tasks including generating one or more recognition results and one or more confidence values for the one or more recognition results, determining whether the one or more confidence values meets a confidence threshold, aborting a remaining portion of the speech recognition tasks for SRS's that have not generated a recognition result, and outputting a final recognition result based on at least one of the generated one or more speech results. 1. A method comprising:receiving, at data processing hardware of a speech recognizer device, a spoken utterance of a user from a telephony server, the telephony server configured to receive the spoken utterance over a network from a user device of the user and route the spoken utterance to the speech recognizer device;initiating, by the data processing hardware, execution of multiple speech recognition tasks in parallel for the spoken utterance, each speech recognition task configured to generate a corresponding recognition result indicating a candidate transcription for the spoken utterance and a corresponding confidence value identifying a probability that the recognition result is correct;determining, by the data processing hardware, whether any of the corresponding confidence values generated by the corresponding speech recognition tasks satisfy a threshold confidence value; andwhen none of the confidence values satisfy the threshold confidence value, requesting, by the data processing hardware, the telephony server to re-route the spoken ...

SPEECH RECOGNITION WITH PARALLEL RECOGNITION TASKS

The subject matter of this specification can be embodied in, among other things, a method that includes receiving an audio signal and initiating speech recognition tasks by a plurality of speech recognition systems (SRS's). Each SRS is configured to generate a recognition result specifying possible speech included in the audio signal and a confidence value indicating a confidence in a correctness of the speech result. The method also includes completing a portion of the speech recognition tasks including generating one or more recognition results and one or more confidence values for the one or more recognition results, determining whether the one or more confidence values meets a confidence threshold, aborting a remaining portion of the speech recognition tasks for SRS's that have not generated a recognition result, and outputting a final recognition result based on at least one of the generated one or more speech results. 120-. (canceled)2128-. (canceled)29. A computer-implemented method comprising:obtaining, by a computer and from a speech recognition system, recognition results and confidence values for a speech recognition task initiated for a received audio signal, wherein the recognition results identify a plurality of candidate representations of the received audio signal and the confidence values identify a corresponding plurality of probabilities that the recognition results are correct;processing, by the computer, the obtained recognition results based on the confidence values for the speech recognition task;generating, by the computer, one or more weighted confidence values for the recognition results based on the recognition results, the confidence values for the speech recognition task, and contextual information related to the recognition results, a value of the one or more weighted confidence values corresponding to a particular recognition result from the recognition results;ranking, by the computer, the processed recognition results based on the ...

Automatic hyperlinking of documents

A system may use a machine-learned model to determine whether to classify a sequence of one or more words within a first document that is being edited as a candidate hyperlink based at least in part on context associated with the first document. In response to classifying the sequence of one or more words as the candidate hyperlink, the system may use the machine-learned model and based at least in part on the sequence of one or more words and the context to determine one or more candidate document to be hyperlinked from the sequence of one or more words. In response to receiving an indication of a second document being selected out of the one or more candidate documents, the system may modify the first document to associate the sequence of one or more words with a hyperlink to the second document.

Business listing search

A method of operating a voice-enabled business directory search system includes prompting a user to provide a type of business and an identifier of a specific business, receiving from the user a speech input having information about the type of business and the identifier, and recognizing, using a speech recognition module, the identifier based on the type of business.

Business listing search

A method of operating a voice-enabled business directory search system includes prompting a user to provide a type of business and an identifier of a specific business, receiving from the user a speech input having information about the type of business and the identifier, and recognizing, using a speech recognition module, the identifier based on the type of business.

Cooperatively training and/or using separate input and subsequent content neural networks for information retrieval

Systems, methods, and computer readable media related to information retrieval. Some implementations are related to training and/or using a relevance model for information retrieval. The relevance model includes an input neural network model and a subsequent content neural network model. The input neural network model and the subsequent content neural network model can be separate, but trained and/or used cooperatively. The input neural network model and the subsequent content neural network model can be “separate” in that separate inputs are applied to the neural network models, and each of the neural network models is used to generate its own feature vector based on its applied input. A comparison of the feature vectors generated based on the separate network models can then be performed, where the comparison indicates relevance of the input applied to the input neural network model to the separate input applied to the subsequent content neural network model.

Business listing search

Generating author vectors

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating author vectors. One of the methods includes obtaining a set of sequences of words, the set of sequences of words comprising a plurality of first sequences of words and, for each first sequence of words, a respective second sequence of words that follows the first sequence of words, wherein each first sequence of words and each second sequence of words has been classified as being authored by a first author; and training a neural network system on the first sequences and the second sequences to determine an author vector for the first author, wherein the author vector characterizes the first author.

Computing device with remote contact list

In one implementation a computer-implemented method includes generating a group of telephone contacts for a first user, wherein the generating includes identifying a second user as a contact of the first user 5 based upon a determination that the second user has at least a threshold email-based association with the first user; and adding the identified second user to the group of telephone contacts for the first user. The method further includes receiving a first request to connect a first telephone device associated with the first user to a second telephone device associated with 10 the second user. The method also includes identifying a contact identifier of the second telephone device using the generated group of telephone contacts for the first user, and initiating a connection between the first telephone device and the second telephone device using the identified contact identifier.

Speech recognition with parallel recognition tasks

【課題】本明細書の主題はとりわけ、音声信号を受け取ること、および複数の音声認識システム(SRS)で音声認識タスクを開始することを含む方法で実施される。 【解決手段】各SRSは、音声信号内に含まれる予想される音声を指定する認識結果と、音声結果の正確さの信頼度を示す信頼値とを生成するように構成される。この方法はまた、1つまたは複数の認識結果、および1つまたは複数の認識結果に関する1つまたは複数の信頼値を生成することを含む音声認識タスクの一部を完了すること、1つまたは複数の信頼値が信頼閾値を満たすかどうかを判定すること、認識結果の生成を完了していないSRSに関する音声認識タスクの残りの部分を停止すること、および生成した1つまたは複数の音声結果のうちの少なくとも1つに基づいて最終的な認識結果を出力することをも含む。 【選択図】図１

Computing device with remote contact lists

In one implementation a computer-implemented method includes generating a group of telephone contacts for a first user, wherein the generating includes identifying a second user as a contact of the first user based upon a determination that the second user has at least a threshold email-based association with the first user; and adding the identified second user to the group of telephone contacts for the first user. The method further includes receiving a first request to connect a first telephone device associated with the first user to a second telephone device associated with the second user. The method also includes identifying a contact identifier of the second telephone device using the generated group of telephone contacts for the first user, and initiating a connection between the first telephone device and the second telephone device using the identified contact identifier.

Business listing search

A voice-enabled business directory search system can be operated by prompting a user to provide a type of business and an identifier of a specific business, and receiving from the user a speech input having information about the type of business and the identifier. Using a speech recognition module, the identifier is recognized based on the type of business.

Computing device with remote contact list

Training and/or using an encoder model to determine responsive action(s) for natural language input

Business or personal listing search

A method of searching a business listing with voice commands includes receiving, over the Internet, from a user terminal, a query spoken by a user, which includes a speech utterance representing a category of merchandize, a speech utterance representing a merchandize item, and a speech utterance representing a geographic location. The method includes recognizing the geographic location with a speech recognition engine based on the speech utterance representing the geographic location, recognizing the category of merchandize with the speech recognition engine based on the speech utterance representing the category of merchandize, recognizing the merchandize item with a speech recognition engine based on the speech utterance representing the merchandize item, searching a business listing for businesses within or near the recognized geographic location to select businesses responsive to the query spoken by the user, and sending to the user terminal information related to at least some of the responsive businesses.

Dynamic barge-in in a speech-responsive system

A method and system for barge-in acknowledgement are disclosed. A prompt is attenuated upon detection of speech. The speech is accepted and the prompt is terminated if the speech corresponds to an allowable response.

Business listing search

A voice-enabled business directory search system can be operated by prompting a user to provide a type of business and an identifier of a specific business, and receiving from the user a speech input having information about the type of business and the identifier. Using a speech recognition module, the identifier is recognized based on the type of business.

Speech recognition with parallel recognition tasks

The subject matter of this specification can be embodied in, among other things, a method that includes receiving an audio signal and initiating speech recognition tasks by a plurality of speech recognition systems (SRS's). Each SRS is configured to generate a recognition result specifying possible speech included in the audio signal and a confidence value indicating a confidence in a correctness of the speech result. The method also includes completing a portion of the speech recognition tasks including generating one or more recognition results and one or more confidence values for the one or more recognition results, determining whether the one or more confidence values meets a confidence threshold, aborting a remaining portion of the speech recognition tasks for SRS's that have not completed generating a recognition result, and outputting a final recognition result based on at least one of the generated one or more speech results.

Cooperatively training and/or using separate input and response neural network models for determining response(s) for electronic communications

Systems, methods, and computer readable media related to determining one or more responses to provide that are responsive to an electronic communication that is generated through interaction with a client computing device. For example, determining one or more responses to provide for presentation to a user as suggestions for inclusion in a reply to an electronic communication sent to the user. Some implementations are related to training and/or using separate input and response neural network models for determining responses for electronic communications. The input neural network model and the response neural network model can be separate, but trained and/or used cooperatively.

Generating author vectors

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating author vectors. One of the methods includes obtaining a set of sequences of words, the set of sequences of words comprising a plurality of first sequences of words and, for each first sequence of words, a respective second sequence of words that follows the first sequence of words, wherein each first sequence of words and each second sequence of words has been classified as being authored by a first author; and training a neural network system on the first sequences and the second sequences to determine an author vector for the first author, wherein the author vector characterizes the first author.

Training encoder model and/or using trained encoder model to determine responsive action(s) for natural language input

Systems, methods, and computer readable media related to: training an encoder model that can be utilized to determine semantic similarity of a natural language textual string to each of one or more additional natural language textual strings (directly and/or indirectly); and/or using a trained encoder model to determine one or more responsive actions to perform in response to a natural language query. The encoder model is a machine learning model, such as a neural network model. In some implementations of training the encoder model, the encoder model is trained as part of a larger network architecture trained based on one or more tasks that are distinct from a “semantic textual similarity” task for which the encoder model can be used.

Recognizing speech in multiple languages

Speech recognition systems may perform the following operations: receiving audio; recognizing the audio using language models for different languages to produce recognition candidates for the audio, where the recognition candidates are associated with corresponding recognition scores; identifying a candidate language for the audio; selecting a recognition candidate based on the recognition scores and the candidate language; and outputting data corresponding to the selected recognition candidate as a recognized version of the audio.

Training encoder model and/or using trained encoder model to determine responsive action(s) for natural language input

Systems, methods, and computer readable media related to: training an encoder model that can be utilized to determine semantic similarity of a natural language textual string to each of one or more additional natural language textual strings (directly and/or indirectly); and/or using a trained encoder model to determine one or more responsive actions to perform in response to a natural language query. The encoder model is a machine learning model, such as a neural network model. In some implementations of training the encoder model, the encoder model is trained as part of a larger network architecture trained based on one or more tasks that are distinct from a “semantic textual similarity” task for which the encoder model can be used.

Selective text prediction for electronic messaging

A computing system is described that includes user interface components configured to receive typed user input; and one or more processors. The one or more processors are configured to: receive, by a computing system and at a first time, a first portion of text typed by a user in an electronic message being edited; predict, based on the first portion of text, a first candidate portion of text to follow the first portion of text; output, for display, the predicted first candidate portion of text for optional selection to append to the first portion of text; determine, at a second time that is after the first time, that the electronic message is directed to a sensitive topic; and responsive to determining that the electronic message is directed to a sensitive topic, refrain from outputting subsequent candidate portions of text for optional selection to append to text in the electronic message.

Speech recognition models based on location indicia

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing speech recognition using models that are based on where, within a building, a speaker makes an utterance are disclosed. The methods, systems, and apparatus include actions of receiving data corresponding to an utterance, and obtaining location indicia for an area within a building where the utterance was spoken. Further actions include selecting one or more models for speech recognition based on the location indicia, wherein each of the selected one or more models is associated with a weight based on the location indicia. Additionally, the actions include generating a composite model using the selected one or more models and the respective weights of the selected one or more models. And the actions also include generating a transcription of the utterance using the composite model.

Cooperatively training and/or using separate input and response neural network models for determining response(s) for electronic communications

Systems, methods, and computer readable media related to determining one or more responses to provide that are responsive to an electronic communication that is generated through interaction with a client computing device. For example, determining one or more responses to provide for presentation to a user as suggestions for inclusion in a reply to an electronic communication sent to the user. Some implementations are related to training and/or using separate input and response neural network models for determining responses for electronic communications. The input neural network model and the response neural network model can be separate, but trained and/or used cooperatively.

Cross-lingual initialization of language models

Cross-lingual initialization of language models

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for initializing language models for automatic speech recognition. In one aspect, a method includes receiving logged speech recognition results from an existing corpus that is specific to a given language and a target context, generating a target corpus by machine-translating the logged speech recognition results from the given language to a different, target language, and estimating a language model that is specific to the different, target language and the same, target context, using the target corpus.

Cross-lingual initialization of language models

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for initializing language models for automatic speech recognition. In one aspect, a method includes receiving logged speech recognition results from an existing corpus that is specific to a given language and a target context, generating a target corpus by machine-translating the logged speech recognition results from the given language to a different, target language, and estimating a language model that is specific to the different, target language and the same, target context, using the target corpus.

Cross-lingual initialization of language models

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for initializing language models for automatic speech recognition. In one aspect, a method includes receiving logged speech recognition results from an existing corpus that is specific to a given language and a target context, generating a target corpus by machine-translating the logged speech recognition results from the given language to a different, target language, and estimating a language model that is specific to the different, target language and the same, target context, using the target corpus.

Cross-lingual initialization of language models

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for initializing language models for automatic speech recognition. In one aspect, a method includes receiving logged speech recognition results from an existing corpus that is specific to a given language and a target context, generating a target corpus by machine-translating the logged speech recognition results from the given language to a different, target language, and estimating a language model that is specific to the different, target language and the same, target context, using the target corpus.

Training encoder model and/or using trained encoder model to determine responsive action(s) for natural language input

Training encoder model and/or using trained encoder model to determine responsive action(s) for natural language input

Systems, methods, and computer readable media related to: training an encoder model that can be utilized to determine semantic similarity of a natural language textual string to each of one or more additional natural language textual strings (directly and/or indirectly); and/or using a trained encoder model to determine one or more responsive actions to perform in response to a natural language query. The encoder model is a machine learning model, such as a neural network model. In some implementations of training the encoder model, the encoder model is trained as part of a larger network architecture trained based on one or more tasks that are distinct from a "semantic textual similarity" task for which the encoder model can be used.

Selective text prediction for electronic messaging

A computing system is described that includes user interface components configured to receive typed user input; and one or more processors. The one or more processors are configured to: receive, by a computing system and at a first time, a first portion of text typed by a user in an electronic message being edited; predict, based on the first portion of text, a first candidate portion of text to follow the first portion of text; output, for display, the predicted first candidate portion of text for optional selection to append to the first portion of text; determine, at a second time that is after the first time, that the electronic message is directed to a sensitive topic; and responsive to determining that the electronic message is directed to a sensitive topic, refrain from outputting subsequent candidate portions of text for optional selection to append to text in the electronic message.

Cooperatively training and/or using separate input and subsequent content neural networks for information retrieval

Systems, methods, and computer readable media related to information retrieval. Some implementations are related to training and/or using a relevance model for information retrieval. The relevance model includes an input neural network model and a subsequent content neural network model. The input neural network model and the subsequent content neural network model can be separate, but trained and/or used cooperatively. The input neural network model and the subsequent content neural network model can be “separate” in that separate inputs are applied to the neural network models, and each of the neural network models is used to generate its own feature vector based on its applied input. A comparison of the feature vectors generated based on the separate network models can then be performed, where the comparison indicates relevance of the input applied to the input neural network model to the separate input applied to the subsequent content neural network model.

Adapting enhanced acoustic models

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for enhancing speech recognition accuracy. In one aspect, a method includes receiving voice queries, obtaining, for one or more of the voice queries, feedback information that references an action taken by a user that submitted the voice query after reviewing a result of the voice query, generating, for the one or more voice queries, a posterior recognition confidence measure that reflects a probability that the voice query was correctly recognized, wherein the posterior recognition confidence measure is generated based at least on the feedback information for the voice query, selecting a subset of the one or more voice queries based on the posterior recognition confidence measures, and adapting an acoustic model using the subset of the voice queries.

Identifying substitute pronunciations

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, including selecting terms; obtaining an expected phonetic transcription of an idealized native speaker of a natural language speaking the terms; receiving audio data corresponding to a particular user speaking the terms in the natural language; obtaining, based on the audio data, an actual phonetic transcription of the particular user speaking the terms in the natural language; aligning the expected phonetic transcription of the idealized native speaker of the natural language with the actual phonetic transcription of the particular user; identifying, based on the aligning, a portion of the expected phonetic transcription that is different than a corresponding portion of the actual phonetic transcription; and based on identifying the portion of the expected phonetic transcription, designating the expected phonetic transcription as a substitute pronunciation for the corresponding portion of the actual phonetic transcription.