Google Inc.

Nathan David Howard

Gabor Simko

Maria Carolina Parada San Martin

Srinivas Vasudevan

Guru Prakash Arumugam

Ramkarthik Kalyanasundaram

Behshad Behzadi

Marcin M Nowak-Przygodzki

Andrei Giurgiu

Quan Wang

Arun Narayanan

Chung-Cheng Chiu

Ishai Rabinovitz

Yanzhang He

Michael Buchanan

Christopher Bo Tandiono

Tom O'Malley

Alex Seungryong Park

James Walker

Maria Carolina Parada San Martn

A method includes receiving a spoken utterance that includes a plurality of words, and generating, using a neural network-based utterance classifier comprising a stack of multiple Long-Short Term Memory (LSTM) layers, a respective textual representation for each word of the of the plurality of words of the spoken utterance. The neural network-based utterance classifier trained on negative training examples of spoken utterances not directed toward an automated assistant server. The method further including determining, using the respective textual representation generated for each word of the plurality of words of the spoken utterance, that the spoken utterance is one of directed toward the automated assistant server or not directed toward the automated assistant server, and when the spoken utterance is directed toward the automated assistant server, generating instructions that cause the automated assistant server to generate a response to the spoken utterance.

Utterance classifier

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for detecting a continued conversation are disclosed. In one aspect, a method includes the actions of receiving first audio data of a first utterance. The actions further include obtaining a first transcription of the first utterance. The actions further include receiving second audio data of a second utterance. The actions further include obtaining a second transcription of the second utterance. The actions further include determining whether the second utterance includes a query directed to a query processing system based on analysis of the second transcription and the first transcription or a response to the first query. The actions further include configuring the data routing component to provide the second transcription of the second utterance to the query processing system as a second query or bypass routing the second transcription.

Detecting conversations with computing devices

A method for automatic speech recognition using joint acoustic echo cancellation, speech enhancement, and voice separation includes receiving, at a contextual frontend processing model, input speech features corresponding to a target utterance. The method also includes receiving, at the contextual frontend processing model, at least one of a reference audio signal, a contextual noise signal including noise prior to the target utterance, or a speaker embedding including voice characteristics of a target speaker that spoke the target utterance. The method further includes processing, using the contextual frontend processing model, the input speech features and the at least one of the reference audio signal, the contextual noise signal, or the speaker embedding vector to generate enhanced speech features.

Joint acoustic echo cancelation, speech enhancement, and voice separation for automatic speech recognition

Detecting continuing conversations with computing devices

Methods, systems, and apparatus, including computer programs encoded on computer storage media for classification using neural networks. One method includes receiving audio data corresponding to an utterance. Obtaining a transcription of the utterance. Generating a representation of the audio data. Generating a representation of the transcription of the utterance. Providing (i) the representation of the audio data and (ii) the representation of the transcription of the utterance to a classifier that, based on a given representation of the audio data and a given representation of the transcription of the utterance, is trained to output an indication of whether the utterance associated with the given representation is likely directed to an automated assistance or is likely not directed to an automated assistant. Receiving, from the classifier, an indication of whether the utterance corresponding to the received audio data is likely directed to the automated assistant or is likely not directed to the automated assistant. Selectively instructing the automated assistant based at least on the indication of whether the utterance corresponding to the received audio data is likely directed to the automated assistant or is likely not directed to the automated assistant.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving audio data encoding an utterance; obtaining an intermediate transcription of the utterance; before a final transcription of the utterance is obtained: i) determining that the intermediate transcription of the utterance is associated with a previously received search query, ii) obtaining one or more results that are identified as responsive to the previously received search query, and iii) storing one or more of the results; obtaining the final transcription of the utterance; determining that the final transcription of the utterance is also associated with the previously received search query; and in response to determining that the final transcription of the utterance is also associated with the previously received search query, providing the stored one or more results for output.

Search result prefetching of voice queries

Growing community of inventors

Mountain View, CA, United States of America

Nathan David Howard